A Field Manual for Railroad Engineers

Survival, Water, Medical Field Manuals

Military Manuals

James C Nagle

Document text

This is a digital copy of a book that was preserved for generations on library shelves before it was carefully scanned by Google as part of a project 
to make the world's books discoverable online. 

It has survived long enough for the copyright to expire and the book to enter the public domain. A public domain book is one that was never subject 
to copyright or whose legal copyright term has expired. Whether a book is in the public domain may vary country to country. Public domain books 
are our gateways to the past, representing a wealth of history, culture and knowledge that's often difficult to discover. 

Marks, notations and other marginalia present in the original volume will appear in this file - a reminder of this book's long journey from the 
publisher to a library and finally to you. 

Usage guidelines 

Google is proud to partner with libraries to digitize public domain materials and make them widely accessible. Public domain books belong to the 
public and we are merely their custodians. Nevertheless, this work is expensive, so in order to keep providing this resource, we have taken steps to 
prevent abuse by commercial parties, including placing technical restrictions on automated querying. 

We also ask that you: 

+ Make non-commercial use of the files We designed Google Book Search for use by individuals, and we request that you use these files for 
personal, non-commercial purposes. 

+ Refrain from automated querying Do not send automated queries of any sort to Google's system: If you are conducting research on machine 
translation, optical character recognition or other areas where access to a large amount of text is helpful, please contact us. We encourage the 
use of public domain materials for these purposes and may be able to help. 

+ Maintain attribution The Google "watermark" you see on each file is essential for informing people about this project and helping them find 
additional materials through Google Book Search. Please do not remove it. 

+ Keep it legal Whatever your use, remember that you are responsible for ensuring that what you are doing is legal. Do not assume that just 
because we believe a book is in the public domain for users in the United States, that the work is also in the public domain for users in other 
countries. Whether a book is still in copyright varies from country to country, and we can't offer guidance on whether any specific use of 
any specific book is allowed. Please do not assume that a book's appearance in Google Book Search means it can be used in any manner 
anywhere in the world. Copyright infringement liability can be quite severe. 

About Google Book Search 

Google's mission is to organize the world's information and to make it universally accessible and useful. Google Book Search helps readers 
discover the world's books while helping authors and publishers reach new audiences. You can search through the full text of this book on the web 



at |http : //books . google . com/ 



of the ^ 

'Unlverdtt^ of Wiecondtn 




Utbrarp 

of the 

xantverdUi^ of Wisconsin 



FIELD-MANUAL 

FOR 

RAILROAD ENGINEERS. 



BY 

J. C. NAGLE, M.A., M.O.E., 

Frofeaaor of Civil Bnoinwring in the Agricultural 
and MechamccU Colleye of lexers. 



SECOND EDITION, REVISED, 
TOTAL ISSUE, ELEVEN THOUSAND. 



; NEW YORK 

; JOHN WILEY & SONS, Inc. 

London: CHAPMAN & HALL, Limited 
1913 




Copyright. 1897, 

BY 

J. C. NAGLE. 



THE 8CICNTIFIC PRCSS 

ROBERT DRUMMONO AND COMPANY 

BROOKLYN, N. Y. ' 



196790 

JUL 20 1915 



<^^f^^^Q 






PREFACE. 



Ea9E of reference and uniformity of notation are essential in a 
book that is to be consulted in the field. With this in mind an 
effort has been made in the following pages to secure a systematic 
arrangement of the subject-matter and uniformity of terms and 
notation. Except for a few cases Greek letters have been avoided 
and a single letter is used to designate an angle. In so far as 
practicable each figure is intended to be self-explanatory, so that 
the explanations necessary in connection with the problems have 
been reduced to a minimum. Algebraic equations stand each in 
a distinct line, thus rendering them more easily read. 

A knowledge of the elements of geometry and trigonometry has 
been assumed, and only in the derivation of a few formulas in 
connection with the theory of transition-curves will any higher 
mathematics be needed. But these formulas may be accepted by 
the reader who is unfamiliar with the calculus without in any 
way affecting his ability to understand their applications or to 
follow subsequent reasoning. 

One can most readily turn to what he wants In a book after hav- 
ing become familiar with its contents in the classroom. Keeping 
this in mind this book has been written so that it may be used as 
a text as well as for reference in the field. Wherever practicable 
solutions to problems have been given in a rigid, general form, 
followed by illustrative examples, so that the student need not 
lose sight of the principle involved while following the solution 
for a particular case. Wherever approximate solutions seemed 
preferable they have also been given and their limitations pointed 
out. 

Free use has been made of the Table of Functions of a One- 
degree Curve, thus reducing the labor of field computations. By 
defining the degree of curve with reference to short chords for 

m 



IV PREFACE, 

sharp curves — and, with tables of Radii, Long Chords, Mid- 
ordinates, etc., based on appropriate equations — the errors result- 
ing from assuming the radius to vary inversely with the degree 
of curve will generally be found to be quite small. 

Chapter I gives briefly the general method of making Re- 
connoissance; Chapter II treats of Preliminary Surveys; while 
Chapter III relates to Location. 

Chapter IV, on Transition-curves, follows the method adopted 
by Professor Crandall, and enables one to locate the transition- 
curve with rigid accuracy where such is necessary. Approximate 
methods are also given by means of which the curve may be as 
easily located as any of the more limited easement curves ordi- 
narily met with. 

Chapter V, on Frogs and Switches, contains all that is necessary 
for their location. The formulas have been arranged to give the 
desired quantities in terms of the frog number whenever the re- 
sulting equations would be easier of application than the trigono- 
metric ones usually given. The turnout tables are unusually full 
and give not only the theoretical lead but the stub lead as well, 
from which the practical lead can be at once found when the 
length of switch-rail is known. 

Chapter VI, on Construction, tells how to set slope-stakes, and 
gives simple methods for computing areas and volumes either 
directly or by the use of tables. A short table of prismoidal 
corrections is given for end sections level, and also a formula for 
three-level sections, by means of which a suitable table ma-y be 
computed if desired. 

The tables at the end of this book have been arranged with a 
view to ease of reference, for, whatever the character of the text, 
the chief value of a field-book must depend upon the ease with 
which the tables may be consulted and upon their extent and 
accuracy. Table IX— Functions of a One-degree Curve— sepa- 
rates the logarithmic functions on the one side from the natural 
functions on the other and will be of assistance in locating these 
tables. Table XVI — Transition-curve Table — reading lengthwise 
of the page, likewise serves to separate the trigonometric tables 
from the miscellaneous tables that follow. 

Some engineers object to the use of logarithmic tables in the 
field, but for them the natural functions are at hand; while for 
those who prefer logarithms the five-place tables of logarithmic 
sines, cosines, etc., will be found .easy to consult and interpolate 
between. 



PREFACE. T 

All trigonometric tables are five -place, and others were carried 
to as many decimal places as their character demanded. 

Tables I, III, IV, and V have been computed to agree with 
the definition of the degree of curve requiring curves sharper 
than 7" to be run with chords less than 100 feet in length, as 
described in the text. Tables XVII and XVIII were also com- 
puted expressly for this book. 

Tables VI and XXVII are from electrotypes from Carhart's 
Field Book for Civil Engineers and were furnished by Qinn & Co. 
Electrotypes of Tables II, X, XII, XIII, XIX, XX, XXIV, XXV, 
XXVI, and also XVI — this last being from Crandall's book, 
T7i6 Transition Curve — were furnished by John Wiley & Sons. 

Of the others, some were arranged from standard tables and 
others adapted in part and extended to increase their usefulness. 

It will be noticed that vertical lines have been omitted wher- 
ever practicable, thus rendering it easier to refer to the tables. 

Acknowledgments are due my associate. Professor D. W. 
Jpence, for aid in making the tabular computations and in reading 
proof. 

J. C. Naqlb. 

CoLLKOB Station, Texas, May, 1897. 



PREFACE TO THE SECOND EDITION. 



In this edition some of the typographical and other minor errors 
that appeared in the first edition have been eliminated. Tables 
XXVIII and XXIX have been added in order to increase the use- 
fulness of the book, and are from electrotypes of tabl. s in Traut- 
wine's Pocket Book, A suggestion has been made by one who 
has had occasion to use the tables quite freely that Table XIX be 
extended so as to give quantities for variations of one tenth of a 
foot in center heights, but such extension would have increased 
the size of the book unduly. When closer approximations are 
wanted than are given by Table XIX the area for the given center 
height can be taken from Table XVII and by entering Table XX 
iirith this as argument the quantity can be at once read off. For 
center heights greater than those given in Table XVII we may 
refer to books devoted exclusively to earthwork computations. 

J. C. N. 

OoLLBGB Station, Texas, January, 1899. 



CONTENTa 



CHAPTER I. 

BSCONKOISSANCB. 
AiKTiOLai i. Objects of Rbconnoissancb— How Mads. 

IKCnON PAOK 

1. Relative Importance of the Work of ReconDoissance and Location.. 1 

2. Object of Reconnoissance S 

3. The Instruments 2 

4. Use of Maps 4 

b. Making the Reconnoissance 4 

CHAPTER II. 

PRELIMINARY SURVEYS. 

Abtxglb 8. Objects; Thb Field Corps; Duties of the Chief. 

(I. Objects of Preliminary Surveys 

7. The Exploration-line 

8. Data Sought in Making Preliminary Surveys 

9. The Field Corps 

10. The Chief of Party, Duties of 



Article S. The Transit Party. 

A. DUTIES OF the MEB1BER8. 

11. Composition of the Transit Party 8- 

12. The Transitman 8 

1&-17. Other Members of the Party g 

18. Instruments 9 

B. TRANSIT ADJUSTMENTS-^THE YERNIER. 

19. Kindof Transit 8 

80. To Ad just the Plate Levels IC 

21. Parallax .• IC 

22. To Adjust the Line of Collimation IC 

28. To Adjust the Standards 11 

vii 



Till CONTENTS, 

8BCTION . PAOK 

84. To Ad j U8t the Level on Telescope o 12 

25. Direct and Retrograde Verniers 13 

2(J. The Least Count of a Vernier o Id 

27. To Read a Vernier 14 

C. ACCESSORIES. 

(1«) The Oradienter, 

28. Description and Method of Using Qradienter 14 

(2?) The Stadia^ or Telemetmr. 

29. Principle of the Stadia 15 

30. Formula for Line of Sight Horizontal 15 

31. Formulas for Line of Sight Inclined 16 

32. The Instrumental Constant, To Find 17 

33. Reducing the Notes 4 17 

D. FIELD-WORK. 

34. station Numbers 18 

85. Hubs or Plugs 18 

36. Reference-points 18 

37. Alignment ..-1 18 

88. Form of Transit Notes 19 

39. Stadia Methods for Preliminary Surveys 19 

B. OBSTACLES IN TANGENT. 

41. To Pass an Obstacle by Means of Parallel Lines. . • 30 

42. To Pass an Obstacle by Angular Deflections 20 

43. To Measure across a River 21 

Article 4. The Level Party. 

44. Make-up and Instruments 23 

45. Work of theLeveler 23 

46. Work of the Rodman 23 

ADJUSTMENTS OF THE LEVEL. 

47. To Adjust the Line of Collimation 23 

48. To Adjust the Level-bubble 24 

49. To Adjnn the Wyes 25 

B. THEORT OF LEVELING. 

jO. True and Aooarent Level 25 

51. The Ersor Due to Curvature 25 

62. The Difference of Elevation of Two Points 26 

C. FIELD-WORK. 

63. The Datum 27 

54 Bench-marks 27 

66. Work in the Field 28 



CONTBIJTa. IX 



8ICCTION * PAGB 

^6. Tbe Level Notes 28 

57. Precautions when UsiDg Level S9 

58. TheRod 29 

Article 5. The Topooraphic Party. 

69. Instruments Used; Area to be Mapped 30 

60. Methods of Recording Data 80 

61. Topographers' Field-sheets 81 

62. Use of the Slope-level 31 

63. Cross section Rods 82 

64. The Transit and Stodia in Topographical Surveying 32 

Article 6. Pbbuminart Estimates. 

66. Map of Preliminary Lines 83 

67. The Profile . ." 88 

68. Preliminary Estimates of Quantities 88 

69. Reportof the Locating Engineer •• 84 



CHAPTER ni. 

LOCATION. 

Article 7. Projecting Location. 

70. Problems Involved in the Paper Location 85 

71 . Hints Regarding Methods of Projecting the Line 85 

72. The Curve-protractor 86 

73. Work in the Field 87 

Article 8. Simple Curves. 
A. definitions and formulas. 

74. Deflnftions 88 

75. To Find the Radius /?, the Degree of Curve Being Known 40 

76. To Find the Length of Curve 42 

77. The Functions of a One-degree Curve 42 

79. To Find D, ^ and C Being Known 48 

80. To Find the Tangent Distance T, I and R Being Known 43 

81. To Find «, Given i and r 44 

82. Given i and A to Find the Long Chord L.C? 44 

83. Ordinates from Chord 45 

84-86. To Find the External .B 48 

87. To Find i?, iff and 1 Given 49 

88. To Find r.JJ audi Given 49 

89. To Find the Deflection Offset from Chord Produced 49 

90. To Find the Tangent Deflection Offset 60 

91. The Sub-tangential Deflection Offset 51 

92. To Find the Tangent Offset z 52 

98. Difference in Length of Arc and Long Chord 53 



Z COKTEKTS. 

B. LOGATINO SIMPLE OURTRS. 
BICTIOir PAGE 

94. To Locate a Curve with the Chain by Offsets from Chords Produced 56 

95. To Locate a Curve by Offsets from Tangent 51 

96. To Locate a Curve by Offsets from a Long Chord 58 

97. To Locate a Curve with Transit and Chain 59 

96. The Index-angle 60 

99. Subdeflection-angles 60 

100-lOL Transit Notes 61 

C. OBSTACLES. 

102. To Pass an Obstacle on a Curve 63 

108. To Locate a Curve when the P. C. is Inaccessible 64 

104. To Pass to Tangent when the P.T. is Inaccessible 67 

106-107. To Pass a Curve through a Given Point 69 

108. To Locate a Tangent to a Curve from an Outside Point 71 

109. To Run a Tangent to Two Curves of Contrary Flexure 72 

D. CHANGE OP LOCATION. _. ^! 

110. To Locate a Curve Parallel to a Qiven Curve 73 

111. To Change P.O. in Order to Make P.T. Fall in a Parallel Tangent. . , 74 

112. To Change R and P.O. to make P.T. Fall in Parallel Tangent, on 

Same Radial Line 75 

113. To Find Change in P.O. or E for a Qiven Change in / 76 

114. Required the Change in P.O. and B for a Qiven Change in /, the 

P. r. Unchanged 77 

115. To Find New Radius for a Qiven Change in T 77 

116. To Find New R to Connect P. C. with a Parallel Tangent 78 

Article 9. Compound Curves. 

A. LOCATION problems. 

117. Qiven Both Tangents and One Radius, to Find the Other Radius ... 80 

118. Qiven One Radius, the Long Chord and the Angles it Makes with 

Tangents, to Find the Other Radius and Central Angles 82 

119. Qiven the Radii and Central Angles, to Find the Tangents, the Long 

Chord,and the Angles it Makes with Tangents. 82 

120. Given the Long Chord and Angles Made with Tangents, to Find 

Both Radii when Common Tangent is Parallel to Long Chord 83 

B. OBSTACLES. 

121. ToLocateSecondBranch when P. C7. is Inaccessible 84 

C. CHANGE OP LOCATION. 

122. To Compound a Simple Curve so P.T. shall Fall in a Parallel Tan- 

gent 85 

128. To Find Change in P.CC, Necessary to Make P.T. Fall in a Par- 
allel Tangen t 86 

124. To Change P.C.C. and Second Radius so P.T. shall Fall in a Par- 
allel Tangent, on Sjsme Radial Line 89 



.1 

CONTENTS. n 



CHMITION PAGK 

125. To Change P.C.C. and Second Radius to Cause P.T. to Fall at a 

New Point in Same Tangent »1 

126. To Substitute a Three-centered Compound Curve for a Simple One. M 

127. To Substitute a Cunre for a Tangent Uniting Two Curves 05 

Article 10. Track Problems. 

188. Reversed Curves, Where to Use 98 

129. To Connect a Located Curve with au Intersecting Tangent 97 

130. To Locate a Y 100. 

131. A Reversed Cur^e between Parallel Tangents 102 

133. A Crossover between Parallel Tracks when a Fixed Length of Tan- 
gent is Inserted 106 

133. A Reversed Curve with Unequal Angles 106 

134. A Reversed Curve between Fixed Points 106 

135. To Connect Two Divergent Tangents by a Reversed Curve 107 

136. To Change P.R.C. so P. T. shall Fall in a Parallel Tangent 106 

187. To Find the Radius of a Curved Track 100 



CHAPTER IV. 
TRANSITION-CURVES. 

Abticlb 11. Theory of the Tbansitiok<;drte. 

188. Elevation of Outer Rail on Curves 110 

139 Requirements of the True Traaisition-curve Ill 

140. Notation Employed ill 

141. Equation of Transition-curve. 112 

142. Transition-curve Angle, / 114 

143. Coordinates of Points 114 

144. Deflection-angles 115 

145. Explanation of Transition-curve Tables 118 

146. To Unite the Branches of a Compound Curve by a Transition- 

curve 119 

147. Length of Transition-curve to be Taken 121 

Article 12. Field-work. 
▲. field formitlas. 

14d. When to Use the Simplified Formulas 122 

149. Simplified Formulas for Transition^urves 122 

150. Offsets l"^ 

151 . Compou nd Curves 1 

B. SETTING OUT TRANSITION-CURVES. 

158. Location by Offsets 1 

154. Location by Deflection -angles 1 

155. Form of Transit Notes for Transition-curves. 1 



XU CONTENTS. 

Abticls 13. Transition curve Problems, 
section paob 

156. Tangent Distances and External for Equal Offsets 139 

157. Tangent Distances, Offsets Unequal 190 

158. Transition-curves Inserted without Changing the Vertex of Cir-. 

cular Curve 131 

159. Transition-curves Inserted with Least Deviation from Old Track.... 133 

160. Transition-curves Inserted at Ends of Long Circular Curve, Cen- 

tral Portion Undisturbed 133 

161. Transition-curve Inserted at P.C.C. by Changing Radius of Second 

Branch 136 

163. To Insert Transition-curves at the Ends of Two Circular Curves 

United by a Common Tangent 188 

163. To Unite a Tangent and Circular Curve when the Offset Cannot be 

Directly Measured 139 

164. Inserting Transition -curves in Old Track 140 

165. Remarks on Tabular Interpolations 140 



CHAPTER V. 

FROGS AND SWITCHES. 
Article 14. Turnouts. 

▲. TURNOUTS from STRAIGHT LINES.? 

166. Definitions 143 

167. To Find the Lead, I, and Radius, /?, in Terms of the Frog Number, 

J\r, and Gauge, g 144 

168. Given R and g, to Find N, i, and Frog-angle, F 146 

169. To Find Theoretic Length of Switch-rail 146 

170. To Find Lead and Number of Crotch-frog for a Double Turnout to 

Opposite Sides of Main Track 147 

171. To Find Turnout Radius and Lead of Crotch-frog in Terms of 

Crotch frog Number 148 

172. To Find Radius of Curve from Point of Middle Frog to Point ot 

Main Frog, Given i^,, JV, and N' 148 

173. Double Turnout to Same Side of Main Track 150 

174. To Find Radius of Curve between Frog-points for a Double Turn- 

out to Same Side of Main Track 151 

175. To Unite Main Track with Siding, Reversing Point Opposite Frog . . 152 

176. To Lay Out a Ladder-track 153 

B. TURNOUTS FROM CURVES. 

177. To Find Lead and Radius for Turnout to Concave Side of Main 

Line ." 154 

178. To Find Lead and Radius, Turnout to Convex Side 157 

1 79. To Find Theoretic I-«ngth of Switch-rail 158 

180. To Unite Main Track with a Concentric Siding 160 



CONTENTS. Xlll 



O. THE rfTUB LEAD. 
SECTION PAGE 

181. Definitions 16*2 

18J. Given iV, <, and p, to Find the Stub Lead 262 

183. Turnout Table and Explanation 163 

184. To Stalce Out a Turnout 165 

185. Curving Ralls 166 

Abticle 15. Crossovers. 

186. Crossover between Parallel Straight Tracks, a Tangent between 

Frog-Rotnts 166 

187. A Cro.«sover in the Form of a Reversed Curve 168 

188. A Crossover with Fixed Length of Intermediate Tangent 168 

189. A Crossover between Curved Main Tracks 168 

Article 16. CROssmo- frogs and Crossimchbufs. 

A. CROS8INO-FROOS. 

191. Length of Rail Intercepted between Two Intersecting Straight 

Tracks 170 

192. Angles of a Set of Crossing frogs, One Track Curved 170 

193. Angles of a Set of Crossing-frogs, Both Tracks Curved 171 

B. CROSSING-SLIPS. 

195. Length and Radii of Slip-rails, Both Tracks Straight 172 

196. Length and Radii of Slip-rails, One Track Curved 172 

197. Length and Radii of Slip-rails, Both Tracks Curved' 173 



CHAPTER VI. 

CONSTRUCTION. 

Article 17. Definitions ; General Considerations ; Vbrtigal 
Curves ; Elevation of Outer Rail. 

<99. The Division Engineer 176 

800. The Resident Engineer 176 

201-204. Definitions 177 

205. To Find the Grade- point, Longitudinal Slope Uniform 178 

206. Vertical Curves iTe 

207 Elevation of Outer Rail on Curves 182 

206. Easing Grade on Curves 183 

Article 18. Earthwork. 

A. setting SLOPE-STAKES. 

209. The Distance Out for Level Sections 186 

210. To Find Position of Slope-stakes for Surface Inclined 184 

211. Cross-section Notes j86 

212. Irregular Sections 187 

U3. Staking Out Openings 187 



XIV CONTENTS. 

BBCnON PAOB 

214. Manner of Marking: Stakes 187 

215. Shrinkage— Growth 187 

216. Borrow-pits, Drainage of , etc 188 

B. ABBAS OF SBCTI0N8. 

218. Area of Three-level Section. 186 

219. Area of Five-level Section 18£ 

220. General Formula for Areas 19C 

221. Explanation of Table of Areas of Level Sections and the Three- 

level Correction 191 

C. VOLUME OP BARTHWOBK. 

228. Where Cross-sections should be Taken 192 

228. Volume by Averaging End Areas 192 

224. ThePrismoidal Formula 193 

fbib. Form of Record 195 

226. The Prismoidal Correction 195 

227. Computation of Volumes when Passing from Cut to Fill 198 

228. Use of Tables of Volumes in Making Preliminary Estimates 199 

229 Side Ditches 199 

230. Earthwork on Curves 199 

231. Overhaul 201 

Article 19. Grade and Ballast Stakes, Cxtlvbrts, Bridgss, 
AND Tunnels. 

232. Grade and Center Stakes 202 

233. Ballast-stakes 202 

285. Openings of , for Culverts. Trestles, etc 1 202 

236. Bridge Piers and Abutments .203 

237. Tunnels 204 

Article 20. Monthly and Final Estimates. 

238. Monthly Estimates 206 

239. Measurements for Earthwork..^ • 206 

240. Classiflcation of Earthwork 206 

241. The Progress Profile 207 

24*2. Masonry Estimates 207 

213. Bridge Estimates 207 

244. Track Material 207 

245. Blank Estimate Sheets 208 

246. Monthly Payments 208 

247. Extras 208 

248. Final Esthnate 206 

249. Acceptance 809 

TABLES. 

Table Showing Length of _TransItioD-curve to be Taken 12l 

Table of Values of ^- Vpt for Stub Lead 163 

Turnout Table 164 

Table of Corrections for Vertical Curves 181 



CONTENTS. X7 

PAGK 

'Table of Elevation of Outer Rail on Curves 182 

Table of Prlsmoidft* Oorrectioiis for Level SectionB 198 

I. Radii of Curves 212 

. II. Minutes in Decimals of a Degree 316 

in. Tangential Offsets 216 

IV. Long Chords and Actual Arcs 217 

V. Mid-ordinates to Long Chords 218 

VI, Logarithms of Numbers.. 220 

VII. Logarithmic Sines and Cosines 238 

VIII. Logarithmic Tangents and Cotangents 253 

IX. Functions of a One-degree Curve 268 

X. Natural Sines and Cosines 296 

XI. Natural Secants and Cosecants 307 

XII. Natural Tangents and Cotangents 320 

xni. Natural Versines and Exsecants : 332 

XIV. Coordinates for Transition-curves 355 

XV. Deflection-angles for TransitionnSurves. ... 356 

XVI. Transition-curve Table 358 

XVIL Area* of Level Sections 371 

XVIII. Corrections for Three-level Ground 375 

XIX. Cubic Yards per 100 ft. in Terms of Center Height 376 

XX. Cubic Yards per 100 ft. in Terms of Sectional Area 382 

XXI. Rise per Mile of Various Grades 386 

XXII. Slopes for Topography 387 

XXIII. Material Required for One Mile of Track 387 

XXIV. Mutual Conversion of Feet and Inches into Meters and Centi- 

meters 388 

XXV. Mutual Conversion of Miles and Kilometers 389 

XXVI. Length of V Arc of LatitudA and Longitude 389 

XXVn. Trigonometric and Miscellaneous Formulas 890 

XXVIII. Square Roots and Cube Roots of Numbers from .1 to 28 895 

XXIX. Squares, Cubes, Square Roots, and Cube Roots, of Numbers 

from! to 1000 896 



FIELD-MANUAL FOR RAILROAD 
ENGINEERS. 



CHAPTER I. 

REC0NN01S8ANCE. 

Article 1. Objects op Reconnoissance— How Made. 

1. The question of the selection of the proper route for n line 
of railway is essentially an economic one, involving not only tlie 
cost of construction, but of maintenance and operation, and a 
consideration of the immediate and future traffic likely to pass 
over the completed road. 

The engineer upon whom devolves the duty of making the 
surveys for a railroad is not often called upon to determine 
whether it should or should not be built, though his prelim iuaiy 
estimate may decide those whose duty it is to do so : the problem 
confronting him is 7iow to secure the best line, answering a gicen 
purpose, for the least cost. Keeping in mind the proper working 
of the completed road, the problem may be divided into two gen- 
eral parts : 

First, The selection of the general route between terminal 
points, and in some cases the selection of the terminals them 
selves. 

Second, The fitting of the line to the ground in such a manner 
as will render the cost of constructing and operating the road a 
minimum. 

The first is by far the more important and difficult operation, 
requiring the highest giade of engineering skill— a fact too sel- 
dom recognized by those selecting engineers for this work. The 
acquirement of the necessary skill can result only from long 
practice and close observation, coupled with the ability to fully 



2 A FIELD-MANUAL FOR RAILROAD ENGINEERa 

grasp and weigh all the complex features of the questioD. A 
passing reference only can be made to it in this little volume, 
which is intended to furnish hints and aids to the better execution 
of the second part. For the beuefit of the beginner who has to 
do with the location and construction a few definitious and hints 
relating to reconuoissance will be given before going on to the 
special problems arising in the work of the railroad engineer. 

2. The Reconnoissance is a rapid, general survey of the area 
through which the proposed railroad must pass, made only with 
such instruments as can be easily carried, and which should ena- 
ble the engineer to restrict the more accurate instrumental work 
that follows to one or two general lines. The time required for 
this part of the work will in general be only a small fraction of 
the time consumed in location, involving the service of very few 
men; yet there is no part of the work more rapidly and im- 
properly done — not always because the engineer in charge under- 
estimates its importance, but because he is not usually allowed 
sufllcient time in which to study thoroughly the area under con 
sideration. 

Properly the reconnoissance includes the determination of the 
terminal points of the road, but the locating engineer is usually 
relieved from the necessity of selecting these points, and the 
question reduces to that of finding the best available line which 
admits of being built, maintained, and operated at the least cost 
betioeen two given points. 

The reconnoissance must be made over an area — not a line or 
lines. Even what seems the most unpromising portion should 
be carefully studied, for the engineer can never be satisfied he 
has selected the best route until he has convinced himself by care- 
ful study that all others are inferior. Too much haste on recon- 
noissance means either a poor line or a much greater expenditure 
of time and money on the preliminary. No amount of notes or 
topography can take the place of an intimate personal knowledge 
of the problems to be encountered, and hence the reconnoissance 
and preliminary survey should be made by the engineer who is 
to locate the road. 

3. The Instruments needed will rarely be more than a pocket- 
compass, hand-level, aneroid barometer, field-glasses, and some- 
times a pedometer or an odometer. 

(a) The Pocket-compass is used to obtain the magnetic bear- 
ings of lines and the angles tlicy inake with each other. 



RECOKNOISSANCB. 9 

(b) The Hand-level enables one to obtain differences of ele- 
vation between points not far apart. 

(e) The Aneroid Barometer gives approximate heights of the 
mercury column, and serves to roughly determine the difference 
of elevation of given points. In addition to the scale giving 
readings in inches, it should have also a scale graduated to give 
readings in feet. If two aneroids, which have been previously 
compared, are read simultaneously, one at each of the points 
whose difference of elevation is desired, or if the same aneroid is 
read at each successively at a short interval of time, during which 
the atmospheric pressure has not sensibly altered, we may find 
the difference of elevation by the formula* 

d = eQm(}og H-iogK)(i + I^±^^), . . (1) 

in which d is the difference of altitude in feet, H and h the 
barometric readings in inches — the logarithms being of the com- 
mon or Briggs kind, 7^ and t the temperatures of the two stations 
in Fahrenheit degrees. 

If the sum of the temperatures, T+t,m taken as 105', formula 
(1 ) reduces to 

ci = 63000(logir-logA) (1') 

Example. — The reading of the barometer at the foot of a 
mountain is 28.8 inches, and at the top 26.7 inches. Required 
the height of the mountain. 

By (1'). d = 63000 (log 28.8 - log 26.7) = 2071 feet. 

The effect of. temperature on the metal of the instrument 
should be considered in the barometric formula when very pre- 
cise work is to be done ; but this correction, being small, may be 
neglected in the rough work of reconnoissance, particularly since 
the makers of the instrument construct it in such a way as to 
compensate, as closely as possible, for such changes of tem- 
perature. 

(d) The Pedometer is an instrument which automatically 
counts the number of steps made by a person when the instru- 
ment is attached to his belt ; then, knowing the average length 
of step, the distance passed over can be readily computed. 

The Odoftieter registers the number of revolutions of a wheel 
to which it is attached, and the number of revolutions multiplied 
Uy the circumference of the wheel gives the space pdlssed over. 

* See Plyinpton*8 Aneroid Barometer, p. 38, for formula (1). 



4 A FIELD- jIAKUAL FOR RAILROAD ENGINEERS. 

4. The Map. — Before beginning the reconnoissance the engi- 
neer should provide himself with the best available map of the 
region to be traversed ; if this is a topographic one, he can at 
once determine from it the lines that are likely to justify an 
examination ; and even if it is only a sketch-map, he can get 
material assistance by observing the courses of the streams and 
remembering that their positions indicate the relative elevations 
of the portion of the region through which they flow. Thus the 
large streams follow the lines of least elevation, and the manner 
in which the lateral streams unite with the principal one indi- 
cates the general trend of the terrain. Two streams flowing 
nearly parallel approacli or recede from each other according as 
the intervening land diminishes or increases in altitude. Two 
streams flowing away from each other on opposite sides of a 
divide, and having their source therein, approach each other 
closest at the point of least elevation, and indicate the position of 
a pass or the lowest point of the dividing ridge. The study of 
any good contour map covering sufficient area will illustrate the 
laws governing the courses followed by streams. 

The elevations of a few correctly mapped points, when obtain- 
able, from the map or otherwise, serve as a guide in tentatively 
fixing on the maximum gradient to be employed and the amount 
of development needed. 

A skillful engineer will thus be enabled to project his lines 
with sufficient accuracy to enable him to select on the ground the 
most feasible route or routes for his preliminaries in the least 
possible time. He should guard against the conviction, however, 
that it is unnecessary for him to look elsewhere than along the 
projected routes ; for the inaccuracies of the map, local peculiari- 
ties, the nature of the excavation and embankment, the number 
and cost of bridges and other mechanical structures, — all these 
may conspire to make the most promising map-line inferior to 
some other whose advantages * have to be sought for on the 
ground. 

6. Having tentatively decided on the limiting grades and cur- 
vature to be employed, the engineer goes carefully over the 
ground, examining the entire area that seems likely to afford 
passage, in order to determine whether a suitable line may be 
secured for the grades and curves previously assumed. With his 
pocket-compass he takes the bearings of lines, and by means of 
the hand-level and aneroid determines differences of elevation. 



RECONNOISSANCE. 5 

Distances are estimated by the eye, paced, and the count taken 
irom the pedometer, or, if the country admits of the use of a 
vehicle, taken from the odometer readings. If a well-gaited 
saddle-horse is used, very good results may be gotten by timing 
him, or by the use of the pedometer if bis stride is uniform. 

But in all cases much dependence must be placed on the ability 
to estimate with the eye differences of elevation and distances. 
The ability to do this with even reasonable accuracy comes only 
from long practice and careful observation, even to the most 
gifted in this uespect. New and unexpected conditions some- 
times deceive even the most practiced eye, but uuder ordinary 
conditions almost any one can train his eye to estimate horizontal 
distances fairly well. Vertical heights are more deceptive, pos- 
aibly because we have less practice in this line, and the mind 
seems naturally to exaggerate the vertical as compared with the 
horizontal ; practice, however, will enable us to make allowance 
for the natural tendency to overestimate heights and slopes. 

The ground should be gone over in both directions, for the ap- 
pearance may be quite different when approached from different 
quarters. Ruling points, such as a pass in the mountains, the 
crossing of a large stream, or a town or city through which the 
road must be built, serve to reduce the problem to a number of 
special ones, each having its own solution. 

In a mountainous region offering a limited number of possible 
routes, but heavy -construction work, it may often happen that 
the location of a line is a much less diflBcult operation than in an 
open, rolling country offering a score of possible lines, between 
which the engineer making the reconnoissance must decide, 
selecting only those that in his judgment seem to justify an 
accurate instrumental survey. 

The engineer must keep constantly in mind all the factors of 
the general problem of economic location and maintenance, and 
successful operation of trains. One line may cost more for con- 
struction and maintenance than another, but less for operation, 
or may invite less traflfic. In all cases, however, the question 
of grades, curvature, length of line, earthwork, and mechanical 
structures are the controlling elements to be considered. 

Having decided upon the route or routes over which to run 
preliminaries, these are marked on the map, and the engineering 
party organized and put in the field, with all the ' necessary 
instruments. 



CHAPTER n. 

PRELIMINARY SURVEYS. 

A RTicLE 2. Objects; The Field Corps ; Duties op the Chief. 

6. The Ol]jects of the preliminary surveys are to secure all the 
data necessary to determine which one of the routes selected on 
reconnoissance is the most feasible, all things considered, and the 
approximate cost of construction. In rough country it will be 
economical to make two, or even three, surveys over the route se- 
lected for location before beginning to place the line in the position 
it is finally to occupy. The first of these is often omitted, and is 
called an " exploration-line " ; it will frequently save the making 
of the more expensive "preliminary" over one or more of the 
routes. 

7. The Exploration-line may be made with either transit or 
compass, and consists of a rapidly run line, made for the purpose 
of determining the maximum curvature and gradients with which 
to project the preliminary. It will not be necessary to make a 
detailed study of the region at this time, the distances and eleva- 
tions, with such sketch topography as may be easily taken, being 
all that is needed. The magnetic bearing of lines is taken by 
the compassman, and the chainmen align each other with the fiag 
set by the flagman. As the progress of the level party will be 
slower than that of the compass party, it will be economical to add 
an extra rodman, and sometimes a recorder. The compassman 
may sketch in the features adjacent to the line while waiting for 
his chainmen, who may be either in front of or behind the com- 
pass. 

The stadia method of surveying — to be spoken of later — would 
seem to offer exceptional advantages for this work — only three or 
four men being needed in addition to the chief. With it, by set- 
ting the transit over alternate stations, very rapid progress may be 
made, and obstacles avoided with as much or greater ease than 
with the compass. 

The exploration -line will more than pay for itself in showing 

6 



PRELIMINARY SURVEYS. 7 

what routes it will be unnecessary to make preliminaries over, 
and in indicating the most feasible one. It should be run over all 
the routes selected on reconnoissance. 

8. The Preliminary Survey follows the exploration, or, when 
this is omitted, comes next after the reconnoissance. It may, with 
advantage, be made |in two parts — first and second preliminary> 
It is made with such instrumental accuracy as the nature of the 
case may demand, sufficient data being obtained to determine the 
best line on which to locate and the approximate cost of construc- 
tion. The rapidity with which this work can be done will depend 
on the care with which the reconnoissance was made. The pre- 
liminary line should approximate, as closely as the eye can deter- 
mine, to the position the located line should occupy, and forms the 
base on which the topographic work rests. In reasonably easy 
country, where exploration-lines have been run, one preliminary 
should suffice for each route, but in difficult regions it will be best 
to run a second preliminary. If portions of the route are easy, fol- 
lowed by difficult parts, it will often be sufficient to " back up '* 
and re-run the difficult portion until a reasonably satisfactory line 
has been obtained. 

9. The Field Corps consists of a chief of party, transitman, 
leveler, rodman, two chainmen, rear rodman or "back- flag," 
stakeman, and two or more axemen. If a topographic pai*ty is 
added, as it should be in any but the easiest country, there will be 
also a topographer with two or more assistants. A cook and 
teamster will be needed with the camp outfit. 

The corps is usually divided into the following parties : 

(a) The Transit Party. 

(6) The Level Party. 

(c) The Topographic Party. 

10. The Chief of Party receives his orders from the chief en- 
gineer, or such other officer as may be in charge, directs the mo- 
tions of the surveying corps, and is responsible for their conduct 
and progress. He provides accommodations and supplies, pays all 
expenses, taking receipts or vouchers for all outlays — ^in dupli- 
cate when required. In the less thickly populated sections he 
must provide tents, wagons, cook, and all necessary camping outfit 
and supplies. He must direct the field operations in person, keep- 
ing in advance of the transit, establish turning-points or hubs, 
and direct the transitman in the proper course. He should keep 



8 A FIELD-MANUAL FOR RAILROAD ENGINEERS. 

a record — or direct the transitinan and topographer to do so— of 
the character of earthwork likely to be encountered, the places 
where drains, culverts, bridges, cattle-guards, etc., are needed; 
the nature of material for embankment, piling, etc., adjacent to 
the line ; the probable amount of clearing and grubbing, and all 
other features likely to affect the cost of construction. He should 
see that the names of property owners and residents along the 
line and the positions and bearings of property lines, when 
possible, are noted. 

He should have authority to discharge assistants — except transit- 
man, leveler, and topographer — whose services are unsatisfactory, 
and in many cases it will be best for him to have entire control, 
engaging or discharging any member of the corps as circumstances 
may require. 

Article 3. The Transit Party. 
A. Duties of the Members. 

11. The Transit Party should consist of a transitman, head 
chainman, rear chainman, rear flagman, stakeman, and as many 
axemen as may be required — rarely less than two even for open 
country. 

12. The Transitman cares for his instrument, keeping it in aa- 
justment; directs the chainmen into line; notes the angle between 
successive tangents as read on plates; notes also the bearings of 
tangents, of highways, streams, and property lines (on location), 
with the plus at which the line crosses them. If there is no 
topographic party he must make sketches, on the right-hand page 
of note-book, of the surface features adjacent to the line; the 
red line down the middle of page represents the transit line, 
whether straight, broken, or curved, to which the sketches are 
adjusted. He must see that the axemen keep in line, in order 
that no unnecessary chopping may be done. Large trees need 
rarely be felled on preliminary, even when a given general course 
has to be followed, for small angles may be turned to avoid them, 
the dellections to right being made to approximately balance those 
to left. 

When the chief of party is absent the transitman is ranking 
man, and will take temporary charge. 

13. The Head Chainman carries a range-pole or "flag," and 
drags the chain, which he must see is straight and horizontal 



PRELIMINARY SURVEYS. 9 

when setting a point for a stake. He directs the stakeman where 
to drive his stake, calling out the number after the rear cliainman 
has read and called out the number on his stake; he keeps the 
axemen in line by setting his flag and going ahead, directing them 
cohere to cut by keeping them in line with the flag and transit. 
The speed of the party is dependent on the rapidity and accuracy 
with which he can set his flag in position, by ranging with stakes 
already set between him and transit, and in seeing that the 
axemen make all their work count. 

14. The Rear Chainman must be careful to hold his end of 
the chain in the proper place, and that it is kept straight and taut 
when the head chainman is setting a stake. He must give all 
pluses, note -the number on each stake as he comes up to it, and 
see that the stakeman has marked it correctly; he must make a 
note of pluses for roads, fences, streams, etc., to be given to the 
transitman later on. 

16. The Stakeman must keep himself supplied with stakes 
about 1|" X 3" X 24", marking the number on them plainly, and 
driving them as directed by the head chainman. 

If sawed stakes are not provided, he must cut the stakes and 
face them for the numbers. He must keep on hand a number of 
plugs or **hubs^" to be driven flush with the ground and having 
the point where flag rested marked with a tack. About ten oi 
twelve inches to the left of and facing the hub a guard stake is 
driven, on which is marked the station number, and which enables 
one to And the hub at any time. 

16. The Axemen do all necessary clearing and chopping in 
order that the transit and level parties may have a clear sightway, 
and yet restrict the work of clearing to a minimum. One of them 
may be detailed to keep the stakeman supplied with stakes. 

17. The Rear Flagman holds his flag on the last turning- 
point for the transitman to use in back-sighting. 

18. The Instruments used by the party are the transit (or 
compass), one-hundred-foot chain or tape, range-poles, and the 
necessary axes and hatchet for axemen and stakeman. 

B. Transit Adjustments — The Vernier. 

19. For railroad work the transit is usually plain, but it is 
often convenient to have a clamp and tangent movement to tele- 



10 A FIELD-MANUAL FOR RAILROAD ENGINEERS. 

scope, a vertical circle, a level on telescope, stadia wires, and a 
gradienter; the solar attachment will rarely be needed. 

20. To Adjust the Plate Levels.— The axis of the instrument 
is set at right angles to the plates by the manufacturer, so that 
when the axis is made vertical the plates will be horizontal. 

In making adjustments remember that a complete reversal 
always doiibles any existing error. 

Place the bubble-tube parallel to a diagonal pair of leveling- 
screws, and bring the bubble to the centre of its run. Revolve 
the instrument 180" on the vertical axis, and the level- tube will 
be parallel to the same pair of leveling-screws as before, but 
reversed. If the bubble has moved from its central position 
bring it half-wAj back by means of the capstan- headed screws at 
the ends of the tube. Relevel and repeat until the bubble remains 
at the centre after reversal. Do the same for the other bubble. 
Both bubbles should remain at the centres of^their tubes during a 
complete reversal. 

21. Parallax is an apparent movement of the cross- wires with 
respect to the object sighted when the eye is moved from side to 
side of the eyepiece, and shows that the image does not fall in the 
plane of the cross-wires. In precise measurements it should be 
removed before making an observation with the telescope. To do 
this, first bring the cross-wires clearly into view when the object, 
glass is turned towards the sky, then, when sighting an object, 
note if there is any relative movement of cross- wires and image 
when the eye is moved from side to side at the eyepiece ; if there 
is, refocus the object-glass until this movement disappears. 

22. To Adjust the Line of Collimaiion is to make the line 
joining the intersection of cross-wires and optical center of objec- 
tive describe a plane perpendicular to the horizontal axis of instru- 
ment. 

First Method. — Level the instrument and clamp the move- 
ments on vertical axis. Sight some well-defined object distant 
about the length of an average sight, and in the same horizontal 
plane as telescope. Reverse the telescope on its horizontal axis, 
and fix a point about as far from instrument as first point, and in 
the same horizontal plane. Revolve the instrument on its vertical 
axis and sight the first point; then reverse the telescope and note 
if line of sight cuts the second point. If not, loosen the capstan- 
headed screws holding cross- wire ring and move the vertical wire 



PRELIMINARY SURVEYS. 

over ons fourth the apparent error — since there were two revei i 
— remembering that the image of the cross-wires is inverted, w i 
that of the object appears in its trne position. Test by repetit : 

SscoND Method. — If the limb graduations can be relied 
they may be nsed in adjusting the vertical wire. With the ins i 
ment level sight a well-defined point, then revolve 180° by vern ( 
plate, reading both verniers; reverse telescope, and note if lint 
sight cuts the point. If not, correct one half the apparent erroi 
moving diaphragm ; then test by repetition. 

The manufacturers adjust the object-glass slide so that the : 
jective travels in the telescope axis, and this adjustment is ;: 
liable to serious derangement. It is well, however, to sometic] 
test by adjusting the line of collimation for both near and distii 
objects. If not correct for both, move the ring which guides 1 
rear end of object-glass slide until the adjustment is correct i[ 
both positions. 

Next make the vertical wire vertical by noting if it coincid 
throughout its length with a plumb-line, or by observing if it c 
vlates from a point, on which the intersection has been fixed, wh 
the telescope is elevated or depressed. Any error is corrected 
turning the ring after slightly loosening the screws holding it. 

The horizontal wire should also be adjusted so that the vdXk 
section of the cross-wires will be in the axis of the telescope ; 
the transit is to be used as a leveling instrument this adjustme 
is essential. 

Drive a stake close to the instrument, and with the telesco 
clamped as nearly horizontal as can be conveniently done reac 
rod held on top of the stake ; about 300 feet distant, and in li 
with first stake and instrument, drive a second stake and read t 
rod on it. Revolve 180** on vertical axis, reverse the telescope a 
bring the horizontal wire to the former reading when the rod 
held on first stake ; if the reading on the second stake is not t 
same as before, correct oiie half the apparent error by moving t 
cross- wire ring. Repeat as a test. The vertical wire should age 
be tested lest the movement of the ring may have caused it 
change. 

23. To Adjust the Standards is to make the plane describ 
by the line of collimation vertical. Set up the transit about as i 
in front of some high building, or other tall object, as the high< 
point that can be sighted is above the base. Level the instrumc 
and fix the intersection of the cross- wires on the highest point tl 



12 k FIELD-MANUAL FOR RAILROAD ENGINEERS. 

can be easily sighted. Depress the telescope and fix a point near 
the base of the building at about the height of the telescope. Un- 
clamp and revolve on the vertical axis until the telescope reversed 
cuts the lower point. Clamp the plates and raise the telescope 
until the cross- wires are at the height of the upper point. If they 
gut it the standards are in adjustment. If they do not, bring 
them half-way back by means of the adjustable screws at the top 
of one of the standards. Repeat ^.s a test. 

24. To Adjust the Level on Telescope is to make the bubble 
stand at the center of its run when the line of sight is horizontal. 
Bring the telescope as nearly horizontal as may be convenient, and 
take readings on the tops of two pegs in the same vertical plane 
with, and equidistant from, the instrument — say 300 feet. The 
difference of readings will equal the difference of elevation of the 
pegs; this difference may be obtained with the wye-level if pre- 
ferred. 

Move the instrument to a point beyond one of the pegs and in 
line with both. Set up as close to nearer peg as convenient, but 
not so close that the rod cannot be easily read. Bring the tele- 
scope as nearly horizontal as possible, and read on both pegs. If 
the difference of readings equals their difference of elevation the 
line of sight is horizontal, and the bubble may be brought to the 
center by means of the adjustable screws attaching the level- tube 
to the telescope. If this is not the case, we must set the telescope 
so the reading on second peg equals the reading on first peg plus 
the difference of elevation ; then read again on first peg and pro- 
ceed as before until the condition is satisfied. Or we may proceed 
as follows : 

In Fig. 1 let the transit be at 0, and A and B be the pegs. AC 
is a horizontal through A, so that CB is the difference of elevation 




of A and B, Suppose line of sight to cut the rods at ^and i>, 
we must find DG bo that the target may be set at the proper read- 



PRELIMINARY SURVEYS. 13 

ing to make the line of sight horizontal. Let 0F= a, FO = 6, 
EA =r, DB = r\ GB = k. Draw DH parallel to GA and 00) 
then JSH=r + k — r', 
From similar triangles 



a4-b 



--m- 



Set the target at a reading GB = QD + r*, sight to G, and the 
line of sight will be horizontal. Bring the bubble to the center of 
its run while the telescope is in this position, and the adjust- 
ment is complete. 

If desired, a correction for the curvature of the earth and re- 
fraction may be introduced, but for short sights this is a useless 
refinement. 

26. The Vernier is an auxiliary scale for measuring smaller 
divisions than those graduated on the limb. There are two 
classes, the direct-reading and the retrograde, according as the 
fractional parts of limb readings are taken on that side of the 
zero of vernier scale towards which the vernier has moved with 
respect to the limb, or the reverse. On the direct vernier a cer- 
tain number of divisions on the vernier equals the same number 
of divisions on the limb, less one ; on the retrograde there is one 
more division on limb than on vernier when the same space is 
covered by both. 

26. The Least Count of a vernier is the smallest subdivision of 
limb graduation that can be read by it, and equals the difference 
of one space on limb and one on vernier. 
Let I = value of one space on limb ; 
V = value of one space on vernier ; 
n = number of spaces on vernier. 
Then for the direct vernier 

nv = (71 — 1)1 ; 

from which we get the least count, 

n 
For the retrograde vernier 

nv = (n-\- ly. 



14 A FIELD-MANUAL FOR RAILROAD ENGINEERS, 
from which the least count is 

the same result as found for the direct vernier. 

So, to find the least count : Divide the value of one limb space h^ 
the number of spaces on the vernier. 

For example : If the limb of a transit is divided to half-degrees 
and the number of spaces on the vernier is 30, the least count 
will be J divided by 30, or ^ of a degree — that is, 1 minute. 

27. To Read a Vernier, take the number of the last division on 
limb back of the vernier zero, then look along the vernier until a 
line is found to coincide with a line on the limb ; add the number 
of this vernier line, multiplied by the least count, to the scale 
reading, and the result will be the required reading. 

C. Accessories, 
(V) The Qradienter. 

28. The Gradienter consists of a tangent-screw having a 
micrometer-head, attached to one of the standards of the transit 
and capable of being clamped to the horizontal axis of the tele- 
scope. It is used — as its name indicates — in running grades, and 
it accurately measures a small vertical angle in terms of its tan- 
gent. The screw is so cut that one revolution moves the tele- 
scope through an angle whose tangent at one hundred feet from 
the instrument has a certain value, usually one foot. The grad- 
uated head is divided into 100 parts, so that one division corre- 
sponds to 0.01 ft. at 100 feet from instrument. 

To run a given gradient, bring the telescope level and read the 
micrometer-head of screw; then turn the screw as many divisions 
as there are hundredths of a foot rise or fall in 100 feet, and with 
a target set at the height of the horizontal axis, points on the 
surface corresponding to the given grade can be found. 

For example : To run a 0.76 per cent grade, move the microm- 
eter milled head 75 graduations from the horizontal. 

When used as a Telemeter, we may either measure the space 
on the rod moved over by the line of sight for a given number of 
revolutions of the screw, or we may note the number of revolu- 
tions required to move the line of sight over a certain space on 
rod. The second method is the more accurate, particularly for 
long sights. 



PRELIMIKARY SURVEYS. 



15 



(2°) The Stadia, or Telemeter. 

29. The Stadia is an instrument for determining the distance 
of a point from the observer by noting the space intercepted on a 
rod by a given visual angle, as determined by two auxiliary wires 
parallel to, and equidistant from, the horizontal wire of the transit 
telescope. When used with an ordinary leveling-rod the wires 
should be adjustable ; if they are fixed (which for some reasons 
is preferable), the rod must be graduated to correspond. In 
addition to the distance of a point from the instrument, the differ- 
ence of elevation is determined by observing the angle made by 
line of sight with the horizontal when the middle horizontal wire 
cuts a point on the rod as high above the ground as is the centre 
of the telescope. 

The horizontal position of the point is determined from its 
magnetic bearing, or the azimuth of line of sight with reference 
to some fixed line, usually the north-south line. 

30. Line of Sight Horizontal.— In Fig. 2 let a and b be the 
stadia wires, AB the intercept on the rod. The secondary axes 

A 




Fio. 2. 



aA and bB pass through the optical center 0, Let h = ab, 
r = AB, d = distance of cross- wires from objective, JD ~ distance 
of rod from objective. 



From similar triangles, 



d 



From optics, 



d "^ d' 



1_ 

'r 



in which /is the focal length of objective. 
Eliminating d from these two equations. 



j>=f+ 



f. 



16 A FIELD-MANUAL FOR RAILROAD ENGINEERS. 



Let c be the mean distance of objective from center of instru- 
ment. Adding this to D gives, for the distance of the rod from 
the center of the instrument, 



l = c+f+ir. 



(2) 



J- may be made constant, when (2) becomes 



1 = a + kr. 



(2) 

31. Line of Sight Inclined. — When the line of sight is not 
level it is difficult to hold the rod perpendicular thereto ; hence 
the rod is held vertical, the angle of inclination measured, and a 
correction applied. In Fig. 8 

Q. 




let r = CD be the reading on rod held vertical ; 

r' = FE, the reading perpendicular to line of sight ; 
// = AO, the horizontal distance from A to B ; 
V= BO, the difference of elevation between A and B ; 
n = BAG, the angle of inclination of line of sight. 

Assume angles AFB and AEB = 90°, from which they rarely 
differ more than 15' to 17'. Then, since FBC = n, 

r' = r cos n. 



PRELIMIKARY SURVEYS. 17 ^ 

Hence AB = a-\- kr cos n. 

From triangle ABG 

H = AB cos n 

.•. H = a cos n-{-kT cos' n (8) 

F=-4-Ssin«.; 
. •, V =■ a^m n-\-kr sin w cos n. 
But 2 sin n cos n = sin Sn. . 
Hence F = a sin n + \kr sin 2» (4) 

32. The Instrumental Constant a [r= c +/ of (3)] maj be 
found by measuring the distance from center of instrument to 
mean position of objective, which equals c ; then focusing on a 
very distant object, preferably a star, and measuring from center 
of objective to plane of cross- wires, which equals/. The sum of 
these distances is a in formulas (3) and (4). 

If the stadia wires are fixed, k may be found by measuring for- 
ward on level ground the distance a from plumb-line, and from 
this point a further distance b ; then note carefully the stadia 
reading r when the telescope is level. Then, remembering (2)', 

a-\'h =za-\-kT. 

. '. A; = — , a constant ratio. 
r 

If the stadia wires are adjustable, we may so adjust k that any 
desired reading may be had for a given length of base. A con- 
venient value of k Is 100, which corresponds to an intercept of 
1 foot on the rod at 100 feet from a point a feet in front of the 
instrument, 2 feet at 200 feet in front, etc. 

33. A Stadia Table based on formulas (8) and (4) is published 
by the D. Van Nostrand Company in Winslow's Stadia Surveying ^ 
and can be used more rapidly than the formulas. Johnson's Re- 
dtiction Diagram, by John Wiley & Sons, gives values of // and V 
graphically. Colby's SHde-rule, manufactured by Mahn & Co., 
St. Louis, gives values of V for distances in feet, yards, or meters 
to tenths of a foot, and can be used with great rapidity. 



18 A FIELD-MANUAL FOR RAILROAD ENGINEERS. 



D. Field-work. 

34. Station Numbers should begin with zero for the ioitial 
stake, and are marked on rear side of stake, from the top down- 
ward, the number of the preliminary, A, B, C, stc, being marked 
on the forward side. The marking should be with kiel, or crayon 
that will withstand the action of sun and rain. Stakes may be 
set every hundred feet or only at even stations, as preferred. 

36. Hubs, or Plugs, are transit turning-points, and are short, 
flat- topped stakes driven into the ground flush witli the surface. 
The flag is held on the top and carefully aligned, the position of 
the point being marked by a tack. A special tack with coucavt 
head offers a foothold for point of flag when used in backsight 
ing.* About 10 inches to the left of and with numbered sid« 
facing the hub is driven a guard-stake to mark its position. 

36. Reference-points are two or more hubs, with guard -stakes 
in each of two lines making a good intersection angle at thb 
point whose position they serve to locate. They should be driver^ 
beyond reach of disturbance, and are used in replacing a dis 
located hub. 

These need rarely be used on preliminary. 

37. Alignment. — It is not intended that the preliminary aniV 
location lines occupy exactly the same position ; hence consider- 
able latitude is allowable in the size and number of angles 
turned, care being taken, however, that the maximum curvature 
need not be exceeded on location. Large trees and other obstruc- 
tions may be avoided by turning a small angle until the obstacle 
has been passed, then m&king a deflection in the opposite sense. 
Bearings of tangents are taken with the needle, to serve as a 
check on the angle read on the plates. 

In easy country not requiring a topographic party large angles 
should not be turned, a succession of small ones with short inter- 
vening tangents being substituted in order to make the prelimi- 
nary profile approximate more closely to the location profile. 
These short tangents may conveniently be the long chords of the 
curve that is to follow. 



* Such a tack is manufactured by the A. S. Aloe Co., St. Louis. 



PRELIN^INARY SURVEYS. 



19 



38. The Transit Notes may be kept in the form below, which 
shows both pages of the note -book. The notes run from the 
bottom up, the right-hand page being reserved for sketches ; the 
red line up the middle of the page represents the transit line, 
whether straight or broken, to which the sketches must be 
adjusted. 



Sta. 


Angle. 


68 




670 


ao^cxL. 


66 




65 




64 




630 


6«2'R. 


62 




61 





Calculated 
Course. 



N. lo 48' W. 



N. 180 12' E. 



Magnetic 
Course. 



N. !• 4^ W. 



N, 18« 15' E. 



Bemarks and Sketches. 



39. Stadia Methods for Preliminary Surveys. — Preliminary 
lines are usually run with the transit, but the compass will 
answer nearly as well in most cases, besides admitting of more 
rapid work. The transit and stadia method might well be em- 
ployed, and would effect considerable saving in the cost of pre- 
liminary surveys. For some reason railroad engineers have not 
regarded it with favor, though it is extensively employed in 
topographic surveying where the map is to be used for work that 
is often more precise than needed for railroad preliminaries. 

Particularly is this method applicable to exploration lines. 
With the transit and stadia the entire surveying corps need not 
exceed five or six men, the instrument man acting as transitman, 
leveler, and topographer all in one. The only objection would 
seem to be in the amount of reduction the notes would need ; 
however, with tables and slide-rule (see 33) this work may be 
very rapidly done. For vertical angles of less than one degree 
the horizontal reduction can be neglected, and with side readings 
for topography the angle may be 5 or 10 degrees without necessi- 
tating the correction. Vertical heights are found by the slide- 
rule or by charts. 

This method would really necessitate the making of a topo- 
p^raphic map along a narrow strip of country, from which the 
profile could readily be taken. With a skilled observer and two 
to four rodmen the progress may be more rapid, and fully as 
good for the purpose intended as the more expensive method 
usually employed, 




20 A FIELD-MANUAL FOR RAILROAD ENGINEERS. 

The transit need only be set at alternate stations (wliicb may be 
any length within the reading limits of the wires), the bearings to 
other stations and points off the line being taken with the needle. 
The horizontal angle should also be read on the plates for points 
on stadia line, as a check on the bearings. 

£. Obstac/es in Tangent 

40. Obstructions to vision and measurement in tangent may be 
avoided in a number of ways, a few of which are given in the 
following problems. Other methods of avoiding them will sug- 
gest themselves in special cases. 

The same devices may be used on location, but it is more im- 
portant to maintain a clear sightway then ; so, when possible, we 
should remove the obstruction. 

41. To Pass an Obstacle by Means of Parallel JUnes.— In 
Fig. 4, is the obstruction, AB the obstructed line. At B set 



E F 


Q H 






r S 








i 


\ i 


3 D 



Fio. 4. 

transit ; turn 90*" and measure BF long enough to clear obstruct 
tion. Set transit at F, make BFO = 90*, and measure FG, 
Move to Cf and backsight to F, making FOC = 90**. Measure 
GC=FB, and move to (7, where the angle OCB is made equal 
to 90"*. CD is the desired line, and BC=FO. 

Otlierwise, at A and B erect perpendiculars; take BF= AB; 
produce EF, and at O and H, beyond 0, erect perpendiculars mak- 
ing G (7 = HI) = FB, CD will be the desired line, and BC = FG. 

42. To Pass an Obstacle by Angular Deflections. 

General Case. Angle anything less tJian 90". 

At B (Fig. 5) on the obstructed line deflect an angle a to one 
side and measure BC, taking C so that after deflecting 2a to the 
other side GD v/i\\ clear the obstruction. Make CD = BC And 
deflect an angle a to the same side as at £; DE will lie in AB 
produced. Draw (7£f perpendicular to BD\ then 

BD = BU+HD = 2BC cosa. , . , . (6) 



PRELIMINARY SURVEYS. 



21 




Fig. 6. 
Example.— Suppose a = W 10', BC=CD = 520 ft. 
BD = 2 X 520 X 0.96959 = 1008.37 feet. 
Special Case. Angle 60 degrees. 

In this case the triangle BDF(Fig, 6) is equilateral and BF = 
BD = DF. 

Should it be inconvenient to run to I> we may stop at C, having 
measured BG, At G deflect 60° and measure GE; at E again de- 




Fio. 6. 
fleet 60° and make EF= BG. At ii^ a final deflection of 60° in the 
opposite sense will put the telescope in the desired line, FO, and 

BF=BG+GE. (5a) 

43. To Pass an Obstruction, such as a River, when the Pre- 
ceding Methods are Inapplicable. 
First Case. Point beyond obstruction visible. 
In Fig. 7 let BG be required. 




Fio. 7. 



At B erect and measure the perpendicular BD ; set instrument 
at D and measure angle BDG = a ; then 

BG=BDtsina (6) 



22 A FIELD-MANUAL FOR RAILROAD ENGINEERS. 

Or, if a trigonometric table is not at band, make GDE-=. 90** and 
fix the point E where DE intersects AB ; measuring EB there 
results-, from similar triangles, 

CB_BD 

BD " EB' 

BD* 
whence CB =-^ (6a) 

Otherwise, if a right angle at B is not convenient, measure 
angles GBD = 6, BJDG = a, and side BB. Then c = 180"- (a+ft). 
From triangle BDC^ 

BC = BD^ (66) 

sm c ^ ' 

Example.— a = 56% J = 70% BD = 400 feet. 

By (66), BG = 400 ^^^. = 409.8 feet. 

' sin 54 

Second Case. Point beyond obstruction invisible. 
At B (Fig. 8) measure angle b and line BE; move to E and 
measure angle p, and set hubs on line EG so the line BG will pass 




Fio. 8. 
between them. Angle z = EGB = 180 -{b + y). Then from tri- 
angle BEG 

BG^BE^^ (7) 

sm 2 ^ ' 

Produce EB to D, where DG will be sure to clear obstruction; 
measure BD, 
From triangle BDG^ 

tan \{a - x) __ BG - BD 
tan^(fi + ic) BG + BD 
But a + a; = &, hence 

tani(«-ir) = ^g-j-|g.teni6. ... (8) 



I^ftELtMIirAftY StftVEtS. J^3 

The sum and difference of a and a? are now known, so both may 
be readily found. 

At D set off the angle a with the transit, and have the chainmen 
stretch a cord between the hubs set on line EC at C. Now signal 
the flagman to move his rod along this cord until the vertical wire 
cuts it at C. Set a hub here and place the transit over it. Sight 
to D or E, reverse telescope and deflect into CH, 

Article 4. — The Level Party. 

44. The Level Party consists generally of two members, the 
leveler and a rodman ; sometimes an axeman is added to keep the 
lodman supplied with pegs for turning-points and in clearing the 
line of sight for the level. As the party follows the transit little 
or no clearing will be needed. The instruments used are a level, 
A rod, and a hand-axe or hatchet. 

46. The Leveler makes all necessary observations with his 
instrument, keeping a neat, accurate record of readings and ele« 
vations ; also the positions and elevations of benches and turning- 
points. He should work out elevations of stations while the rod- 
man is going from one station to the next ; he mast see that the 
rodman gives him readings at points where the longitudinal slope 
changes suddenly, recording the plus. He must plot his profile 
at night, or at such times as the chief of party is likely to need it. 
The rodman 's readings at turning-points should be checked. 

46. The Rodman holds his rod at each station, calling out the 
number. If stakes are set only at even stalious, be must hold his 
rod midway between stakes, tbe point being found by pacing tbe 
distance. Target-readings need only be taken at turning-points 
and benches, and tbe rodman should keep n record of these in 
his "peg-book," checking the calculations of leveler for heights 
of instrument and elevations of turning-points. At any marked 
surface change he will hold his rod, calling out the plus to leveler. 
He must assist the leveler In plotting up the notes. 

A. Adjustments of the Le¥eL 

47. To Adjust the Line of Collimation Is to bring the inter- 
section of the cross-wires into the optica! axis of tbe telescope. 

Set up and level the instrument, then bring the vertical wire 
into coincidence with a plumb line or vertical edge of a building, 



24 A FIELD-MAKtJAL FOH RAILROAD ENGINEERS. 

at the mean length of sight, and note if the verlicjil wire is truly 
parallel thereto. If it is not, loosen the cupsttuj- headed screws 
holding cross wire ring and turn slightly so that the wire is 
parallel to the vertical line. 

Loosen the wye-clips and bring the vertical wire into coin- 
cidence with the line and. clamp the instrument. Rotate the 
telescope in the wyes 180** and note if the wire coincides with the 
line. If not, correct one half the error by loosening one and 
tightening the opposite, of the capstan-headed screws that hold 
the cross-wire riug in place, remembering that the image of 
the cross wires is inverted by the eyepiece. 

Turn the telescope until the horizontal wire is parallel to the 
plumb-line or edge of building, and make the same lest and 
correction. Repeat for both wires. The horizontal wire is the 
one on which the accuracy of leveling depends, but it is wise to 
have both adjusted. Their intersection should remain on a point 
during a complete rotation of the telescope in the wyes. 

48. To Adjust the Level-bubble is to bring the axis of the 
level- tube into the same vertical plane with the line of coUimation, 
and to make the bubble stand at the center when the line of sight 
is horizontal. 

Since the axis of the telescope coincides with the line joining 
the center of the wye-rings (which requires these to be of the 
same size), it is sufficient to make the axis of the bubble parallel 
to this line. 

{a) With the telescope over one diagonal pair of leveling- 
screws and the clips loosened, bring the bubble to the center of 
its run ; then turn the telescope, in the wyes, a little to either side 
of the vertical plane through the telescope and note if the bubble 
remains at the center. If not, correct the error by means of the 
screw at end of the level-tube case arranged for b\teral movement, 
iiepeat until the tube may be rotated half an inch or more to 
either side of vertical without movement of the bubble. Thia 
adjustment is made merely to prevent error from failure to set 
level-tube vertically beneath telescope. 

{b) With the wye-clips opened well out, again bring the bubble 
to the center of its run ; remove the telescope from wyes and 
turn it end for end, then carefully replace it in the wyes. Should 
the bubble fail to remain at the center, bring it half -way back by 
raising the lower or depressing the higher end of tube at the 
points of attachment to telescope. Relevel and repeat as a test. 



PRELIMINARY SURVEYS. 



25 



49. To A<^ust the Wyes is to make the axis of the telescope 
perpendicular to the vertictil axis. With the wye-clips closed 
place the telescope over one pair of leveling-screws aud bring the 
bubble to the center of its run ; then turn the telescope half-way 
rouud ou its vertical axis, so that its euds have changed places. 
If there is any error,. correct by bringing the bubble Jialfway back 
to center by means of the screws connecting wyes with level-bar. 
Repeat until the bubble remains in the center during a complete 
revolution. 

B. Theory of Leveling, 

50. When the level has been adjusted the line of collimation 
N\\\ describe a plane parallel to the horizontal plane tangent to 
he earth's surface at the point where the instrument is placed. 

Ji. level surface, such as the surface of still water, will coincide 
with this plane only at the point of tangency, and will depart 
farther and farther therefrom as the point considered recedes 
from the instrument. For short sights this difference may be 
neglected in railroad work, as will presently be shown, but for 
long sights a correction must be applied. 

The effect of curvature is to make objects appear lower than 
rhey really are, while the refraction of a beam of light, due to 
tlie greater density of the layers of air nearest the earth's surface, 
has a contrary effect. Experience shows the average error due 
to refraction to be about one seventh of that due to curvature. 



51. The Error due to Curvature at any point is the deviation 

of a tangent line from true level, as j^ p 

the point recedes from the point of " 
tangency. 

Let be the center of the earth, T 
the point of tangency, and iVthe point 
where the error due to curvature is 
desired. Let the notation be as shown 
in Fig. 9. From the right triangle 
OTP, we have 




From which 



Fia. 9. 



t^ 



2i2 -f c * 
Now, since c is always very small compared with 2i2, the 
quotient resulting from the division of f^ by 2R will not differ 



26 A FIELD-MANUAL FOR RAILROAD ENGINEERS. 



n 1 « 



3 <« 
'1 R' 



sensibly from that obtAined by dividiug by 2B + e. Therefore 
we write 

.= 4. ........ w 

For t = 1 mile, B = 8963 miles, c = about 8 inches. Hence 
for any other distance in miles we have, for c, 

c = 8 Xt' inches (9a) 

The correction for refraction is about -^^j, hence we have, 
from (9), 

or, closely enough, 

C = .85c (10) 

Example —What is the correction for a half-mile sight? 
For one eighth of a mile? 

By (9a), c = 8 X HY = 2" for first case, 

and c = 8 X (J)' = 0".125 for second case. 

By (10) the final correction is 

c = 0.85 X 3 = 1".7 for first case, 

c = 0.85 X 0.125 = 0.106" for second case. 

52. The Difference of Elevation between two points not so 
far apart but that a rod may be read on each from some inter- 
mediate point may be readily found from these rod-readings. 

In Fig. 10 let the instrument be at i, A and B the points 
whose difference of elevation is desired. Let r = AD, r' — BC. 
Since the line of sight, DC, is horizontal, the difference of 






Fio. 10. 



elevation will evidently be r* — r. When the distance from 
/ to -4 equals tliat from I to B the errors due to curvature 
evidently balance. 



PRELIMINARY SURVETS. 



27 



When the points are so situated that the rod cannot be read 
on both from one intermediate position of the instrument, an 



D 




_. F 




A 


J 




C 


^^ 




r 




4 


^^^ 


-^ 


ij^^^^5^ ^^ 



Fia. 11. 

auxiliary point or points must be used and readings taken on 
tliese points in pairs. Thus in Fig. 11 suppose the difference 
of elevation of ^ and 5 required : 

With the instrument ""at / read on A and some intermediate 
point E, Cousideriug the backsights as plus and foresights as 
minus, the difference of elevation of A and E is AB — FE, 

Again, with the instrument at /' the difference of elevation of E 
and B is QE-^ CB, The sum of these differences equals the dif- 
ference of elevation of A and B, and may be written {AD -|- OE) 
— {EF-{- CB), or, in general, the sum of t7ie backsighU less the sum 
of the foresights equals the difference of elevation, 

C. Field-work, 

63. A Datum is a level surface so taken that it shall lie below 
the lowest point likely to be reached by the profile, to which the 
surface elevations are referred. It is often spoken of as the 
datum-line or datum-plane, and is the zero of elevations. 

64. A Bench-mark is a permanent mark, such as a copper or 
other bolt let into the top of a solidly fixed stone, whose height 
above the datum is known; it may be simply a mark on a stone, 
OH a tack driven into the projecting root of a tree, upon which 
the rod may be read. In any case it must be so situated that it 
cannot change its elevation nor is likely to be disturbed within 
the time for which it is intended to be used as a standard of 
reference. 

The elevation should be marked on some object adjacent to 
the bench, with the letters B. M. indicating the nature of the 
point. 



28 A FIELD-MANUAL FOU RAILROAD ENGINEERS, 

56. The Field-work consists in finding the elevation of a 
number of points on the line established by transit party suffi- 
cient to give, when plotted, a fairly correct outline of the surface 
as seen in profile. 

A bench-mark is taken at the beginning of the line, and its dis- 
tance above mean sea-level or other datum is known or assumed. 
The level is set with one pair of leveling screws in the line to be 
run (in order that any change in the position of the bubble may 
be easily corrected), and the rod is read on the bench. This read- 
ing plus the elevation of bench gives the height of instrument 
{H, I.) above the datum. 

Readings are taken at every hundred feet along the line, or 
oftener if the surface changes greatly, until a point is reached 
beyond which it is desired to move the level. A peg is driven 
firmly into the ground and the rod i-ead on this ; the height of 
instrument less the rod reading will give its elevation, as it will 
for the intermediate points. This poipt is a temporary bench, 
and is called a turning-point. It should be marked by a guard- 
stake if it is desired to use it again. The instrument is now car- 
ried beyond the turning-point, set up, and the whole process 
repeated. Benches and turning points should be read to hun- 
dredths or thousandths of a foot, intermediate points to tenths. 
Turning-points are marked or T. P. in the notes, and their 
positions, as also the bench-marks, noted by both leveler and 
rodman in their note-books. 

56. The Level Notes may be kept in any convenient form 
that is easily understood. The following is used more exten- 
sively, perhaps, than any other: 



Sta. 


B.S. 


H.I. 


F.S. 


Elev. 


Remarks. 


B.M. 


5.618 


ii05.613 


.... 


200.0 


j B. xM. on root of L. O. tree ec to 
1 right of line. 









2.8 


203.8 




1 






0.8 


204.8 




2 






5.7 


199 9 




8 






7.8 


197.8 




4 






99 


195.7 







L120 


196.810 


10.4;>3 


195.190 


On peg at 4-1-80' - 20^ to left of line, 
by small P. 0. tree. 


5 






6.d 


190.0 




6 






4.5 


191.8 





Here the elevation of the datum was taken 200. 00 feet below 
the first bench-mark. The instrument was set up near Station 2, 



PRELIMINARY SURVEYS. 29 

iod a reading of 5.613 taken on the bench; this was written in 
the B. 8. column, and when added to the elevation of the bench 
gives the height of instrument, 205.613. A reading of 2.3 was 
taken on Sta. 0, recorded in the F.8. column, and when sub- 
tracted from the H.L yields an elevation of 203.3. The eleva- 
tions of other points were determined in the same way. A little 
beyond Station 4 the rodman drove a peg and held the rod on it, 
yichling a reading of 10.423 and an elevation of 195.190. The 
instrument was then moved to a point near Station 7 and a read- 
ing of 1.120 taken on the peg; this added to 195.190 made the 
new H. L 196.310, and the process continued with this H, L 

In most cases it will be sufficient to read benches and turning- 
points to hundredths and intermediate points to tenths. 

It will be seen from the notes that any error in a turning point 
causes the same error in all succeeding points. To guard against 
this the rodman is required to keep a " peg-book," in which the 
heights of instrument and elevations of turning-points are re- 
corded, and which must check with the leveler's record. 

67. Wind and sunshine affect the accuracy of the work with 
tlie level, as is also the case with the transit. For very great 
accuracy a calm, cloudy day is the best, but the railroad engineer 
<»annot always choose the best times for his work, and must take 
^uch precautions as may be possible while he exercises the great- 
est care to prevent and detect errors. The adjustments should 
l)e tested at least once a week, even when the greatest care has 
been taken, for unequal expansion and other causes may con- 
spire to cause them to change. 

By making foresights and backsights to turning-points about 
equal the error due to curvature will be eliminated; the readings 
of rodman at these points should also be checked. The rodman 
should hold his rod vertical, which is sometimes accomplished 
by meims of a level attached to rod; or the leveler can tell by his 
vertical wire when the rod is in the same vertical plane with the 
instrument, and by causing the rodman to wave his rod back and 
forth slowly, after clamping the target, he can tell if the hori- 
zontal wire just bisects the target at its highest position. 

68. The Rod should be graduated to feet and tenths, reading 
by target at turning-points and benches; intermediate readings 
are made by the leveler at his instrument. Strength and dura- 
bility are essential qualities. The Philadelphia rod seeou to 



30 A FIELD-MAKUAL FOR RAILROAD ENGINEERS. 

answer tlie purpose as well as any other now manufactured: the 
Troy rod may be used in the same manner as the Philadelphia 
rod, but is lighter and less able to stand rough usage. 

Article 5. The Topographic ParTIt. 

59. The Topographic Party follows the level and secures all 
the data necessary for making an accurate contour-map of a strip 
of country extending as far each side of the preliminary as may 
be needed for the intelligent projection of the location -line. 
This distance may vary from 50 to 300 or 400 feet, its width de- 
pending on the difficulties to be encountered and the degree of 
precision with which the preliminary approximates to the final 
location-line. The lateral slope of surface is obtained at the 
stations of preliminary by means of the hand-level and tape, by 
the slope-level or clinometer, by cross section rods, or by the 
transit and stadia. Strictly speaking the topography includes all 
the surface features, but for railroad work the surface elevations, 
streams, and nature of surface are the most imporUint; it may 
be necessary to note the positions of roads, buildings, etc., and 
should always be done when practicable without undue loss of 
time. A pocket-compass will be of use in observing the bear- 
ings of lines. 

60. There are two methods of recording the data obtained; 
one by means of notes and sketches in a book, the other by 
drawing the contours directly on the field sheet as ihe data are 
obtained. Station elevations can be taken direct from the leveler's 
notes, and constitute the base on which the contour elevaliotife 
rest. 

Suppose the hand-level to be used and the notes kept in a book, 
to be afterwards transferred to the map. Starting with the 
known center elevation, the topographer notes the height of his 
eye above the ground and calculates the height of center above 
or below the next contour; from this the reading of the rod when 
held on this contour is found, being the height of station above 
contour plus the height of eye. He directs the slopeman in or 
out on a line at right angles to preliminary until this reading is 
given by the hand-level; the distance out is then measured and 
recorded, just as in setting slope-stakes, and the slopeman di- 
rected into position on the next contour, in the same manner. 

Thus if 5-foot contour.intervals are employed, and the station 



1 



PRELIMINARY SURVEYS. 



31 



cievaliou is 321.6 feet ftnd the height of eye 5.3 feet, we shall have 
fw the reading at ihe 320-fo()t contour 5.3 + (321.6 - 320)= 6.9. 
Motion the slopeman down the slope until his rod reads 6.9 and 
measure the distance out, suppose 21 feet. The 315-foot contour 
will be 5 feet lower, giving a reading of 11.9, which may be 
found in like manner at, siiy, 80 feet out. As the rod reads only 
to about 12 feet the lo^wgrapher must move out to this last point, 
and with the reading 5.3+5= 10.3 find the 310-foot contour in 
ihe same way. On the up-hill side the 325-foot contour will be 
found with a reading of 5.3 - (325 — 321.6)= 1.9 feet, and other 
contours in like manner. 
The notes may be written thus 



Sta. 


Left. 


Center Elev. 


Right 


824 


305 310 315 320 
193' 125* 80 ' 21 


321.6 


825 830 335 340 

27* 56* -80' 112 



The number above the line is the contour elevation, the num- 
ber below its distance out from center. 

If preferred the elevation can be taken at regular distances out 
and recorded as above; the position of the contour will then be 
found by interpolation when mapping the work. 

61. If the topography is to be plotted in as the work progresses 
the topographer must have a light drawing-board with a pocket 
and flap on buck for holding the sheets on which the transit-line 
has been plotted the night before ; the station elevations are 
marked on the line and the contour positions spotted in as ob- 
tained by slopemen, after which the contours are sketched in. 
Points where contours cross transit-line are found in the same 
manner as side points. The size of the sheets will depend on the 
taste of topographer and size of drawing-board; 17x24 to 19x28 
inches are good sizes. 

The topographer will soon learn to guess at the position his 
contours will occupy at the next station ahead, and will sketch 
them in lightly, to be erased and corrected when necessary. It is 
often sufficient to take lateral readings at every second or third 
station. 

62, If the Slope level is used, the inclination of the surface is 
obcained; then by th^ use of a scale constructed to show the 



32 A FIELD-MANUAL FOR RAILROAD ENGINEERS. 

horizoDtal distance apart of contours, for tbe given contourin- 
terval, for slopes varying from 1* to 20**, the position of contours 
can at ouce be spotted on the map. Wellington recommends the 
use of t!)e altazimuth as permitting the employment of either 
method at ^ill— the altazimuth being merely a haud-levcl with 
a clinometer attached. 

63. Oross-section Rods are measuring- rods 10 or 12 feet long 
carrying a level-bubble. By placing one end at the center, 
bringing the rod horizontal, and noting the, height of the end of 
rod on the down-hill side, the slope may readily be obtained and 
the contours worked in as before. For very rough, broken 
ground this method may be preferable to either of the othera. 

64. If the Transit and Stadia are employed, very elaborate 
topography may be taken with very little fieki work, but the ob 
servations require considerable reduction. With a suitable topo 

. graphic protractor and the slide-rule mentioned In 33, the large 
number of points that may be obtained from each setting of the 
transit may be readily plotted and their elevations marked on the 
plot, after which the contour-lines can be worked in, and other 
features mapped. For small vertical angles no horizontal reduc^ 
tion is needed. 

While not generally favored by railroad engineers in the past, 
this method is probably the most rapid and economical of any so 
far employed in topographic work. 

Article 6. Preliminary Estimates. 

66. After completing the field-work of the preliminary survey 
the party is usually disbanded, only the transitman, leveler, and 
topographer being retained to assist the chief of party to complete 
the map, profile, and estimate of cost. 

66. The Map may be drawn to any suitable scale, but less than 
400 feet to the inch is not to be recommended where it must be 
used in projecting location. The transit-line is laid down first 
and the topography worked in afterwards from the field-map or 
topographer's notes. If it is wanted on a continuous sheet, the 
transit- line must first be drawn on a succession of small sheets, 
which are added as the plotting progresses, a new sheet being 
slipped under the edge of the preceding and tacked down whea 



PRELIMINARY SURVEYS. 33 

required. The overlapping edge is marked by a number of short 
lines extending over onto the sheet beneath, to enable one to re- 
place in the proper position. When the line has been plotted the 
sheets are pasted together and the whole shifted so as to bring the 
transit-line over the continuous sheet. Angular points are then 
pricked through and the line drawn on the continuous sheet. 
Ordinarily it will answer to have the map drawn on a succession 
of small- sheets, to be joined together as required. 

The plotting had best be done by bearings, though it may be 
Jone from the deflection angles, provided care is used to check 
frequently by bearings. Otherwise an error in one angle will 
throw all the remaining portion of the line out of position. 

If more than one preliminary was run, they should all be shown 
on the same sheet whenever possible. 

67. The Profile will be drawn by the leveler on profile-paper, 
and shows a develoi^ed vertical projection of the line. The scale 
will depend on the paper used. There are three scales in general 
use. styled respectively Plates ** A," "B," and "C." There is 
also a metric profile-paper. Plate * * A " has the vertical exagger- 
ated 20 to 1 as compared with the horizontal and is the best to 
use where much rockwork is expected. The vertical exaggera- 
tion of Plate '• B" is less than of Plate "A"; this plate is most 
used for ordinary eartliwork. 

A strip of color laid on below the surface-line, and fading out 
at the lower edge, adds greatly to the appearance of the profile. 
The tentative grade- line and points of change should be drawn in 
red. 

68. Preliminary Estimates of quantities are made by assuming 
a grade-line and drawing it on the profile; then the cuts and fills 
are taken from the profile, and the corresponding quantities ob- 
tained from Table XIX for the base the road is intended to have 
when completed. The nature of the work, whether ordinary 
earth or rock, can. of course, be only roughly estimated. 

Bridging is estimated from the profile where piling or framed 
bents may be used, but where piers and long spans are needed 
special surveys with soundings are required, Culverts, drains, 
cattle guards, cross-ties, and rails for main line and sidings, 
switcli stands, buildings, right of way, clearing, and other factors 
entering into the question of cost must all be considered and 
allowed for in making up the estimate. 



34 A FIELD-MANUAL FOR RAILKOAD ENGINEERS. 

EDgiDeering expenses and unforeseen outlays that are sure to 
arise should have a liberal allowance. 

69. The Report of the chief of party should set forth the ad- 
vantages and probable cost of each of the several lines run 
when there is more than one. Ou this report frequently depends 
whether or not the line is to be located, and it should be clear 
and exhaustive, though plainly and concisely worded. The map 
and profile form an integral part of the report and show from 
what data the estimates were derived. 



CHAPTER in. 

LOCATION, 

Article 7. Projecting Location. 

70. After the preliminary has been mapped and the topography 
worked in, the engineer proceeds to make a paj^er location for his 
guidance in the tield. The solution of the varied and complex 
problems that confront him are more or less interdependent. 
The guiding principle, applicable to all departments of engineer- 
ing, that the best structure is tJuit which for the least cost best an- 
swers the purpose fai* which it was intended, should control, even 
thougii the resulting structure be inferior, in point of scientific 
design, to some other. The best road as regards construction and 
grades may be a failure because of excessive first cost, while 
the cheapest construclion will entail such heavy operating ex- 
penses that it may be equally unprofitable. The alignment must 
be as free from curves as possible, while heavy grades are at the^ 
same time excluded; these two requirements conflict and must 
be as well adjusted as possible. The amount of earthwork, of 
bridging and other structures must be kept down to the lowest 
limits. 

71. Starting at the' summit of the most difficult portion of the 
route, assume a starting-point and elevation; with the dividers set 
at such a distance to the scale of the map as will give a fall of one 
contour space— or half space— for the assumed grade, step down 
the slope in such a way that the dividers fal^ each time on the 
next lower contour, or half-space, according to the fall assumed in 
setting dividers. If curve compensation is allowed, the dividers 
must be reset for each curve, for the same fall, since the grade 
will be slackened on curves. The points at which the dividers 
fall are lightly spotted on the map ami connected by a grade 
contour, which represents the surface line having the required 
gradient. This line will be too broken to be used as a location- 

35 



o6 A FIELD-MANUAL FOR RAILROAD ENGINEERS. 

line, so we have then to draw on the map a succession of curves 
aud tangents that will approximate sufficiently close to it. at the 
same time that a proper balance is maintained between earthwork 
and curvature. 

Having lightly plotted the proposed line, the elevations are 
transferred to profile- paper, thus giving a profile of the line. 
With a fine thread stretched along the profile, to represent the 
grade-line, adjust the cuts and fills to suit the nature of the work. 
Ill general, fills are cheaper than cuts both in construction and 
maintenance; and especially is this true where a shallow surface 
luyer of earth is underlaid by rock. It may happen that the 
material from excavation must be used in embankment, whieii 
the cuts aud fills must be made to balance by shifting the gi*ade- 
line until this appears to be the case on the profile. 

At the stream crossings the grade-line must be kept safely 
above high-water mark, so that sufficient waterway is provided, 
and allowance made therefor. 

After locating the most difficult portions pass on to the easier 
work, returning later on to study the effect this will have on the 
part first located. It may be necessary to go over the projection 
several times before you can be reasonably sure that the best loca- 
tion has been projected; even then the study of the line in the 
field will cause many of the details to be altered, sometimes 
materially; 

Long grades are to be preferred to short ones, but questions of 
economy may necessitate the latter in order to lighten work; care 
must be taken that the grades are not so badly *' chopped" that 
they interfere with the easy riding of the train. 

In projecting the line it will generally be best to strike the 
curves first and draw the tangents afterwards, though it some- 
limes happens that long tangents will control the curves; when 
tin's is the case the tangents are drawn to intersection and the 
c'.irves afterwards put in. 

When transition-curves are employed, a slight offset should be 
made at the beginning and end of curves to allow for their inser- 
tion in the field. These offsets will be so small that it is ijsetess 
to attempt to show them to scale. 

72. A Curve-protractor will be of material assistance in find- 
ing the degree of curve required to unite two tangents that have 
been laid down on the map. It consists of a transparent, semi- 
circular protractor having a series of cuiTes from 30' up to 8** 



LOCATION. 37 

plainly cut upon it. The curves are on both sides^ those on the 
reverse side having their concavities turned iu an opposite sense 
from those on the face. The scale is usually 400 feet to the inch, 
and in any case the map and protractor must be drawn to the 
same scale. Sometimes a set of cardboard or hard-rubber curves 
are used, but they are inferior to the curve-protractor. To use 
it. simply prolong tangents to intersection 'and theu place the 
protractor so that the curve admitting of the best grade is tan- 
gent to the two straight lines. Mark the points of tangency, 
which will be the beginning and end of curve. When the curve 
is required to pass through a given point the proper curve may 
be immediately found by trial, whereas the calculations would 
require some little lime. 

Reversed curves should never be allowed on main lines. Suffi- 
cient taugent should be interposed to allow space for easing off 
the superelevation of outside rails, or for the insertion of tran- 
sition-curves when these are to be employed. 

73. The Field Corps is substantially that required on the pre- 
liminary survey, and the methods of work pretty much the same, 
except Ihut curves must now be run iu, and this necessitates more 
clearing. If first and second location- lines are to be run (and it is 
real economy to run both), it will not be necessary to have the 
stationing continuous on the first, so the pluses arising from 
** backing up" need only be noted and eliminated when the final 
location-line is run. If transition -curves nro to be inserted, they 
need not be run the first time, the proper offset being made at 
vhe P. T. or P. C of the circular curves, which latter are to be run. 

On the final location-line the stationing must be continuous, 
beginning with zerp. The stakes are marked as on the pre- 
liminary survey, and all hubs that are likely to be used again 
must be referenced in, the reference-hubs being set well out of 
the way of disturbance by the plow or scraper. 

The leveler should make bench-marks every 1000 or 2000 feet, 
to be used in running check-levels and in giving grades later on. 

From the paper location the notes should be made up in the 
office, to serve as a guide in the field; however, noatteuipt should 
be nfade to adhere rigidly to them, since slight errors in the 
mapping will affect the projected line, while in the field the line 
may be shifted here and there so as to fit the ground more snugly 
and accord more closely with what the nature of the earthwork 
demands. 



38 A FIELD-MANUAL FOR HAILROAD ENGlKEEftS. 

The highest skill of the eogineer is required to secure the best 
locatioD-line, and he should have all the time he needs. Uiidue 
haste on location — as on recouuoissance and prellmlnaTy — is 
almost sure to result in increased cost of construction. 



Abticle 8. Simple Curves. 

A, Definitions and Formulas, 

74. The Circular Curves that are usually employed to unite 
straight reaches of the railroad may be simple, compound, or re- 
versed. The use of reversed curves should, however, be limited 
to turnouts and cross overs. 

a. A Simple Curve is the arc of a circle. 

b. A Compound Curve consists of two simple curves, of differ- 
ent radii, both on the same side of n common taugeut. 

c. A Reversed Curve is made up of two curves of contrary 
flexure having the same or different radii, and a common tangent. 

d. The Point of Curve (P. C.) is the end of tangent and begin 
ning of curve, as at A, Fig. 12. 




Fio. 12. 

6. The Point of Tangent (P.T.) is the end of curve and be- 
ginning of tangent, as at B of Fig. 12. 

/. The Point of Intersection (P.I.) is the point where the 
tangent at the P. (7. and P.T. intersect when produced. {D of 
Fig. 12.) 

g. The Intersection Angle {D is the angle at the PL be- 
tween the tangents meeting there, and equals the angle at the 
center. 

h. The Tangent Distance (T) is the length of the produced 
tangent measured from the P. C. or P.T. to the P.I. The term 



LOCATION. 



39 



tangetit is applied to a»y straiglit portion of the lioe, but the letter 
T will be used to designate the produced portion only. 

i. The Mid-ordinate (if) is the portion of the radius inter 
cepted between the arc and chord when it cuts the chord at its 
middle point. 

j. The External (E) is the part of the radius produced to the 
P.I.y intercepted between curve and the P.l. 

k. The Long Chord {L,G.) is the chord joining the P,C. and 
P. T, Frequently the term is applied to any chord longer than 
the unit chord. 

L The Radias will be denoted by R, 

m. The Point of Compound Cunre (P. G. G. ) is the point of 
common tangency of the two branches of a compound curve. 
(See Fig. 13.) 



/' 


\p.c.c. 


p.l. 


^ 






-^ 


>. 


> 


> 


y^t 



P.R.C. 



Fio. 18. 



n. The Point of Reversed Curve (P.R,G.) is the point of 
common tangency of the two branches of a reversed curve. 

o. The Degree of Curve (Z>) is the angle at the center sub- 
tended by the unit chord. In the United States this chord is 100 
feet, in England 66 feet, and where the metric system is em- 
ployed it is taken at 20 meters. Any convenient chord length 
may be taken, but for uniformity American engineers have 
adopted the chord of 100 feet, and unless otherwise stated it is 
always so understood when we speak of the degree of curve. 

Half the degree of curve is called the deflection-angle, since 
it is the angle to be deflected from the tangent to the chord. 

If there were any practical method of measuring around tJie 
curve instead of along the chord, an accurate and convenient 
ratio for expressing the radius in terms of the degree would be 
had. Thus if D is the angle at the center subtended by the arc 
of unit length, we have, where a is this unit arc, 



40 A FIELD-MANUAL FOR RAILROAD ENGINEERS. 



2nR= a—jj' 



Hence 



i? = 



a 360 



27t D' 

When a equals 100 ft. this becomes 
100 860 



i? = 



2;r 



(11) 



(11') 



R varies inversely as Z), so that knowing the radius for a V 
curve, we should have only to divide this by D to get the radius 
for a D** curve. 

Since the clioi*d is employed instead of the arc, we determine 
R by means of the following problem 

75. Oiven the Chord (7, and Degree of Curve 2), to Find the 
Radius R. 

In Fig. 14, AB \^ the chord (7, OE a pei-pendicular from the 
center upon AB 




From the right triangle AEO we have 
R sin \D = IC. 



Whence 
When (7 is too ft., 



R = -ri^ = iCcosec 4Z). 
sm \I> ^ ■ 



50 



R = - — rr, = 50 qos^c ID. 
sm ID * 



(13) 



(m 



LOCATION, 41 

Comparing results given by formula (12') with those given by 
ill'), we have for a few curves: 

De^rree of Curve. R by (12'). R by (11'). Difference. 

1 5729.65 5729.58 0.07 

2 2864.93 2864.79 0.14 

8 1910.08 1909.86 0.22 

5 1146.28 1145.92 0.36 

7 819.02 818.51 0.51 

10 573.69 572.96 0.73 

14 410.28 409.26 1.02 

20 287.94 286.48 1.46 

The difference is seen to be about one half a foot for a 7** 
curve, one foot for a 14" curve, and one and one-half feet for a 
20** curve. 

Up to a 7** curve the difference is inconsiderable, and we may 
stake out curves with 100-foot chords. From 7 to 14 degrees 50- 
foot chords may be used. Therefore 

^="slp=^^''^'^''*^- • • • • • (12'«> 

For curves from 14** to 28** we should use 25-foot chorda, 
for which 

i? = ^^ = 12.5cosecJ2). .... (m) 

Above 28" shorter chords— say 10 feet— should be used, if the 
curve cannot be struck from the center. In this case 

Table I of radii was computed by formulas (12'), (12'a), and 
(12'&). 

In practice it is customary to take tbe radius of a 1** curve ns 
5730 feet and to assume the radii to vary inversely as the degree ; 
thus for a 4" curve the radius would be Ii=: ^jifl = 1432.5 feet, 
while by Table I it is 1432.69 feet— a difference of only .19 foot ; 
for a 12" curve R=m^ = 477.5 feet, while by Table I it is 
477.68 feet. 'I'he effect of taking 5730 instead of 5729.65 for the 
radius of a 1" curve is to reduce the error resulting from the 
assumption that It equals 5730 divided by the degree of curve. 



42 A ^-iKLD-MANt/AL J?Oft RAILROAD EKQlKEEIlS. 

76. The Length of Ounre (L) is found by dividing the angle 
at the center (wliich equals the intersection angle) by the deirree 
of curve, the result being in chains and decimals of a chain. The 
number of P.C,-{-L will give the station number of P. T, 

Example.— The P. C, of a 4* curve having / = 26** 30' is at sta. 
104 + 12.5. Find L and the number of the P. T, Here 

X = ?^ = 6.625 chahis. 

104.125 + 6.625 = 110.75 ; hence the number of P. T. is 
110 + 75. 

77. Use of the Table of Functions of a One-degree Curve. — 
In the location of railway curves geometrical accuracy will 
frequently be of less importance than rapidity of lield-work, so 
long as errors are kept within certain limits. 

On tangents slight errors of alignment may readily be detected 
by the unaided eye, but on curves these are not so apparent. 
Moreover it is not likely that the trackmen will keep them up in 
the exact position of their location. 

To simplify and shorten the field computations engineers make 
use of a table of functions of a V curve, and assume these func- 
tions for other curves to vary invereely as their degree, or directly 
as their radii. Table IX gives values of the tangent distances, 
long chords, mid-ordinates, and externals for a V curve, the 
radius of which is taken as 5730 feet. To find these functions 
for other curves, divide the tabular values by the degree of curve. 
The error resulting from this assumption will, in any practical 
case, amount to no more than a few tenths or hundredths of a 
foot. 

Table IX may also be used as a metric curve table, the tabular 
values being taken as meters instead of feet. If the unit metric 
chord is 20 meters long, this may be taken as one fifth of the 
tabular unit chord; so to use the table multiply the metric degree 
by 5 and enter the table with the result as a value of 2). 

For instance, a 2° metric curve having / = 40" would have a 

mid -ordinate equal to ^ — -^ = 34.56 meters. 
-© X o 

For the approximate radius of a metric curve divide 5730 by 5 

times the degree. Thus a 4° metric curve would have i?= -^ =. 

4X6 



LOCATION. 43 

:=: 386.5 metei's. For the exact mdius make use of formula (12). 
Thus for a 4° curve having 20-meter chords R = — — ^ = 286.54 

meters, a difference of ouly .04 meters. 

If a metric curve is to be retraced with a 100-ft. chain, we 
convert the metric degree to the degi-ee referred to 100-ft. chords 
by the relation that a 100-ft. chain = 1.524 chains of 20 meters 
each; a 20-meter chain = 65.618 ft.; one foot = 0.3048 meters; 
one meter = 3.2809 ft. 

It will sometimes be a sufficiently close approximation to lake 
the 20 meter chain as two thirds of a 100-ft. chain; this will make 
the metric curve nearly two thirds of the degree the same curve 
would have when laid out with a 100-ft. chain, and the curve with 
106-ft. chord? nearly three halves of the degree as laid out with 
the 20-meter chain. Thus a 4° metric curve would be equivalent 
to a 6' curve laid out with a 100-ft. chain. 

In the problems that follow two methods of solution will be 
given when practicable—the first being rigid, while the second 
As based on the use of Table IX. To shorten the formulas the 
subscript 1 will be written after the letters T, L,C., M, and E 
when these are the functions of a 1° curve. Thus Tx "4- 28' means 
r.he tangent distance for a 1** curve when /= 28**, L.Cx ^ 16** 
/;he long chord for a 1° curve when /= 16", etc. 

78. Tables of Natural and Logarithmic Circular Functions. — 
Many engineers prefer to work altogether by tables of natural 
sines, cosines, etc., and time may often be saved by their use. 
Nevertheless logarithmic tables are of frequent advantage, even in 
the field, and the more important ones, such as Ihe logarithmic 
sines, cosines, tangents, and cotangents, together with the loga- 
rithms of numbers, are given in the back of the book along with 
the tables of natural functions. 

79. Given R and G to Find Z>. 

From equation (12), 

smW = i^ (13) 

80. Given / and R (or D) to Find T. 

If D is given, find R by (12); then in Fig. 15 from triangle 
OAB we get 

y=i?tani7. (14) 



44 A FIELD-MANUAL FOR RAILROAD ENGINEERS. 



By Tablb IX.— Find the tabular value of T for the given 
angle /; then 

T 

(14a) 




Example.— /= SS** 40', Z) = 4"; required T, 
By (14). r = 1433.69 tau \T 50' = 460.91 feet. 

By (14a). r= -^'^ 



= 460.85 feet, a result differing from the 



value found by the rigid method by only 0.06 foot. 



81. Given I and 7" to Find i? or D 

From (14), 

T 



R = 



tan il 



= TcoHl 



Then by Table I the degree inay be found. 
By Table IX. 



(15) 



(15a) 



82. Given /and D to Find the Long Chord L.C. 
First find R by (12) or (12'), or by Table I ; then from the 
triangle OAF of Fig. 15, 

AF =R sin il. 
.-. AG = 2AF= L. C, = 2R sin ^7. . . . (16) 



LOCATION. 



45 



By Table IX. — Find the tabular L, (7. for the given angle /; 
then 



z.a = 






(16a) 



83. Given the Radios B and any Chord C to Find the 
Ordinate to the Curve at any Point. 

First Method. —In Fig. 16 let HE be the chord C\ HK= a 
and KE = 6, the segments into which it is divided by the ordi- 




nate y. Draw the radius through K\ call the portion between 
chord and curve /. By geometry, 



from which 



ah 

y = 



2.R^y'' 

But y^ is small compared with 2i?, and hence we write 
ah 



^■ 



"2i?* 



(«) 



Now y does not differ sensibly from y' in the cases met with in 
pnictice, so we write 

ah 



y = 



2/2' 



(6) 



46 A FIELD-MANUAL FOR RAILROAD ENGINEERS. 

If we write B = -yr-, formula {b) becomes 

abP 

^^ Too ~ *"' 100 ~ ^' *^^ substitute in (c), giving 



or very nearly 



y = lmnD (17) 



y is given in feet when m and n are in chains and decimals of 
a chain. 
At the mid-point Fym^n, and y = M. 

.-. M^in'D (18; 

Caution.— Formulas (17) and (18), while very convenient for 
field use in passing obstructions, are liable to error when very 
long chords or large values of D are used, since they give results 
that are too small. 

If we write the arcs HN^ NB for a and &, we shall get i-esults 
that are too large, yet about as near the true values as by taking 
m and n to be the segments of the chord. To illustrate we will 
find a few values of M and compare with the true values taken 
from Table V. 



Degree Length Mid-ord. Mid-ord. Mid-ord. 

of of by by by 

Curve. Arc. M=l(HF)^D, M=yiHQ)*D. Table V. 

2 2 Stations. 1.75 1.75 1.75 

2 6 *' 15.69 15.75 15.69 

5 2 '* 4.37 4.38 4.36 

5 6 " 38.51 39.38 39.06 

8 2 ** 6.96 7.00 6.97 

8 4 *' 27.29 28.00 27.75 

8 5 " 42.02 48.75 43.20 

8 6 *' 59.43 63.00 61.98 

From this it appears we may use formula (18)— and (17) as 

well — taking either the segments of the arc or chord for curves 
not exceeding 4" wilh arcs up lo 600 ft. ; for curves from 4' to 6" 



LOCATlOHr. 47 

they may be used up to 600-ft. arcs, while for cui-ves between 
6** and 8** not more ihau 400 feet of arc may be taken. 

Srcond Method.— First determine the mid-ordinate. In 
triangle OEF, 



0F= VH" -id 
then 



M= FO = R - i^B" " iG' (19) 

To find ordinate -4(7 distant d from the mid-point ot EH, draw 
OB=d parallel to HE; draw AB at right angles to HE. Then 



BA = i/B* - d*. 
Therefore 



C^ = y= Vi? - d»- i/iJ^-JCT*. . . . (20) 

Third Method.— If the chord C is short, we may regard the 
arc as an arc of a parabola, for which it is known that ordi- 
nates vary as the product of the segments into which they divide 
the chord. The mid-ordinate being known, we have 

y __ af> 
M - (J60« 

.obM 
.-. y=4-g5- (21) 

From formula (h) we have for y = lf, a = 5 = ^C, 

^- -2^-85 (^) 

The mid-ordinate for any other chord C is 





M, 


= 85- 


Hence 








Jfi 


C"« 




M 


"" (7»* 




.\ Mx-- 


iC'V 


If C" = 


: \C. this gives 


\cj 




J/, 


= iM. . 



48 A FIELD-MANUAL FOR RAILROAD ENGINEERS. 

This last relation affords an easy method of staking out n curve 
when the niid-ordinate of a given chord has been determined. 
Fust erect the ordinate if at the mid point of the chord; then 
join the ends of chord with the extremity of the ordinate just 
measured; the lengths of these chords do not differ much from 
JC; at Iheir mid-points erect ordinates equal to i-Sf, giving points 
on the curve. Proceed in like manner for other points until a 
sufficient number have been located. 



84. Given R and / to Find the XSxternal B, 
In Fig. n S=GB=OB- 00, 
But OB = R sec J/ and 00 = i?. 

.*. i? = i2(secii~l) = i2ex8ec J/. . . . (24) 

By Table IX.— Find E for a 1" curve for an intersection 
angle /; then 

E, 



E = 



D' 



(24a) 



85. Given T and / to Find E. , 

In Fig. 17 draw BC perpendicular to AB, and produce AO to 

b/ 




intersect BCt^i C. BCis parnllel to AO, and the triangles AOO 
and CGB&re similar; hence BC = BO = E. In the right triangle 
ABC, angle BAG^\BAF= \L Therefore 

(25) 



E= rtan J/. . . 
EzERCiBS.— Derive equaitiou (25) from (24). 



LOCATION. 



49 



86. Given if and /to Find ^. 
From trigonometry, 

sec J/ = 
Insert this iu (24) and we get 



cos J/* 



E=R 



1 ~ cos J7 
cos 4/ 



(«) 



But from Fig. 17. if = i2(l - cos \I), Substitute in (a) : 
M 



E-. 



cos \I 



= M sec J/. 



87. Given jSTand /to Find R 
From (24), 

B E 



i? = - 



sec J/ — 1 ex sec J/ 

88. Given /and i? to Find /. 
From (25), 

E 



^E 



cos J/ 
vers \I' 



T = 



= J^cot J/ 



(27) 



(28 



tan J/ 

89. Given the Chord G and Degree of Curve D to Find 
the Chord Deflection Offset d, 

Iu Fig. 18 extend EA to if, making AH = EA = ^45; join 




FiQ. 18. 
£r and B and draw ^-fiT to the mid point of HB. Then 
HK= EB= Csin 1/). 
.-. d=2 HB = 2(7 sin i/) 



(29) 



50. A FIELD-MANUAL FOR ICAILUOAD EXGINEEUS. 

When = 100', 

(f = 200 sin |2) (29') 

\n 
If we write sin i-Z> = ^ from (12) in formula (29), there results 

<« = -^- • . . (30) 

For curves up to T» G = 100'; hence 

^^-R-' • • (80) 

For curves from 7° to 14'', C = 50'; therefore 

^ = ^ (80") 

For R write -^-, and (30'), for G = 100, becomes 

.= SZ, = ,.««„ ,8,, 

and for G = 50, (30") becomes 

d = J-^2> = .43632> = .873.g.. . . . (81') 

Example.— Find d for a 6** curve, G = 100 feet. 
By (29'), cf = 200 X 0.05234 = 10.47 feet. 

By (80'), d = g^ = 10.47 feet. 

By (81), d = 1.745 X 6 = 10.47 feet. 

90. Given the Chord G and Degree of Curve D to Find the 
Tangential Deflection Offset /. 

Ill Fig. 18 make EF (tangent at E) equal to EA, and join F 
with A. Draw EO to tlie mid-point of FA. Angle AEO = 
GEF = \D\ hence, from the figure, 

AG=zQF= G&iniD. 
\ t = 2GsiulD, (82) 



LOCATION. 51 

When G = 100 feet, 

< = 200 siu JD (33') 

Since JD is small, we may write, without material error, 

in 

sin iD = i sin ^D; then, writing sin JZ) = ^, as in 89, we get 

' = M (^> 

Making G = 100 ft. aud writing R = ^ gives 

When C = 50 feet, (33) yields 

t = .218Z> = .436 X ^. ..... (33") 

Example.— Find ^ for a 6** curve, 'C = 100 ft. 
By (320 i = 200 sin 1^ 30' = 5.24 ft. 

By (33'), = .873 X 6 = 5.24 ft. 

91. To Find the Subtangential Deflection Ofiset t' for a 
Subchord G' 

First Method.— By formula (13) fiud the angle at the center 
subtended by the subchord G'\ call this angle U, From (32), 

i' = 2C7'sinii)' (34) 

Second Method.— In Fig. 19, wiih Ens center strike the arcs 
FO and AH, taking EF - G' and 
EA = G ; prolong EO to B. Now 
assuming tbat the chords G' and G 
are proportional to their central 
Angles we have 

AB t / N 

-C-^G" ' • (^). 

From the similar sectors EFO Fio. 19. 

and EAB, since EB = C, 

Ji_G' 

AW^ a 




W 



52 A FIELD-MANUAL FOR RAILROAD ENGINEERS. 
Multiplying (a) and (b) together, term by term. 



Whence 






-(?)■• 



f=i[^\ (35) 

ExAMPiiB.--Find t' for a 7' curve when (7 = 60 ft. 



Here 

By (34), 
By (32'), 

By (35), 



fU) 
ly = ^ X 7* (very nearly) = 4** 12'. 

^ = 2 X 60 X 0.01832 = 2.20 ft. 
t = 6.11 ft. 



92. To Find the Tangent OflFset z. 

In Fig. 20. EB=e is the required offset. Let AS = n chains = 
lOOn feet. AE=FB, the half-chord 
having the mid-ordinate AF = EB ; 
hence we have, by formula (18), 

2 = |»«2). . . . (86) 

In this formula we may take n to 
be either the length of AE or the arc 
AB, in chains. If taken equal to AE 
the offsets will be slightly too small, 
while if taken equal to AB they will 
be a little too large. The use of the 
formula is limited to small values of 
n and D, as was pointed out in 83. 
(See Caution.) 
Formula (36) is easy of application and of frequent use in 

locating curves by offsets from the tangents. For curves up to 

4" n may be as great as 3, but for sharper curves it should 

be less. 
Example.— Find six offsets to a 4" curve at points 50 ft. apart. 

measured around the curve. 




FiG.ao. 



LOCATION. 53 

jBy successive applications of (36) we have 

for» = l, 2 = JxiX4= 0.88 feet 

n = l, « = J X 1X4= 8 50 " 

» = f « = i X f X 4 = 7.88 " 

n = 2, « = jx4x4 = 14,00 " 

n = i « = JXV-X4 = 21.88 " 

n = 3, « = JX9X4 = 31.60 " 

The last value of z is in error by about 0.2 ft., but for setting 
stakes on construction this difference is not material so long as 
the alignment beyond this point does not depend on it.. In 
setting track-centers the completed road-bed is available and the 
stakes may be set with the transit, in the usual way. 

93. l^ifiference in Length of a Circular Arc and its Long 
Chord. 
First Method. — Let the central angle be a degrees. By (13), 

Changing degrees to circular measure, a (in ;r meas.) = — 

180 

t= ^, The length of arc is /2a = ^^. Then 

_• 
Arc — chord = B-—^ —c (37) 

07. o 

Second Method. — An easy approximation may be found as 
follows : 

Referring to Fig. 17. AE = c, GF= M. Let -4(? = ft = ^ 4- aj. 

From the right triangle AFQ 

From which a? = — ; — (a) 



64 A FIELD-MANUAL FOR RAILROAD ENGINEERS. 

NeglectiDg the x in deDOiuiuator as small compared with e 
gives 

*=T <*> 

Alfl 

Then will 2ft-c = 2a» = =- (88) 

From Huygens' approximation to the length of a circular arc 

(see WUliamsou's Differential Calculus, p. 66), arc = — 5 — . 

8 

Therefore 

Arc — chord = —^ c = 4(2* — 0). . . (0) 

Inserting the value of 25 — c from (38) gives 

Arc — chord = -g— (d) 

When the arc is not very great we may write c = 100;ii , where 
ni is the number of chains contained in the arc AE. From (18), 
remembering that ni = 2n, 

if=0.218ni«I^. 

Inserting these values of c and if in (<f). 

8 (.2l8)*ni^i ?* ^ _1_ 
3 lOOw, " 800" 

Example.— Find the difference in length of arc and chord of 
a 4** curve when ni = 6 stations. 

The central angle is 4x6 = 24°; then, from Table IV, 
c = 595.74. 

By (37), 

24 
Arc - chord = 1482.7 X ^ - 595.74 = 4.84 ft 

07. «5 



By (39), 
Ar< 

Remark.— Formula (38) is interesting as showing what a com- 



. , , 6 X 6 X 6 X4X4 . . . . 

Arc - chord = ^^r = 4.82 ft 



LOCATION. 



55 



paratively small iucrease in leugtb of llmt is caused by a consid- 
erable latei-al deflection in alignment. For instance, a lateral 
deflectiou of 2000 feet is made at the mid-point of a line 40,000 
feet long ; what will be the increase in length? 

By (38) the increase is ^^^' = 200 feet, giving for the 



increased length 40,200 feet. 



40,000 



B. Locating Simple Cijrves, 

94. To Locate a Curve with the Chain by Ofiiidts from 
Chords Produced. 
In Fig. 21 let the P. C, fall at B. If BC is a full chain, prolong 




Fio. 81. 



the tangent AB to IT, making Bff= BC; HO will equal t, which 
may be calculnted by (32) or (33'). With B as center, strike an 
arc with radius BIT, and with H as center and t as mdius strike 
an arc , at G, where these arcs intersect, set a stake. Produce 
BC to JSr, making CK-BG-CD\ strike the arc KD from C as 
center ; make the chord KB = d, calculated from (29'), (30'), or 
(31). Set a stake at D and proceed in like manner for the other 
points until the P.T. is reached, where FPis made equal to t 

Usually the P C. does not fall at a full station ; then EC = t\ 
which may be found by (34) or (35). Using this value of t\ we 
locate C as above. At B make BR = i'. and prolong BG to 
L ; make LD = t and set a stake at D. EM will equal d, and 
may be located as before. 

We may regai-d KD as equal to KL + 1, and, finding, KL, 



^ 56 A FIELU-MAKUAL FOR BAILROAD EKGIKEERS. 

measure KD and set D without locating R. To do this we have 
the similar triangles BHC and CKL, from which 

CK BC 
and therefore since KG = CD, 

In like manner at F we have 

PN=t^, and J^P=</ 
hence 

Make EQ = ti\ prolong QF, and we have the tangent at F. 
ExAMPLB.— Given the P.C. of a 5"* curve at 106 + 20 and the 
angle of intersection 22"*, to locate the curve. 

22 
Here £ = -r- = 4.4 stations. 

5 

Therefore the number of the P. T, is 

106.20 4- 4.4 = sta. 110 + 60. 

BCiu this case is 80 ft., and by (33') 

t = 0.873 X 5 = 4.37 ft. 

By (35), r = 4.37 X f^)'= 2.80 ft. 

Set off HC = 2.80 ft., and at D make 

KD = 2.80 X ^ + 4.37 = 7.87 ft. 

At E make ME = d = 8.72 by (31). This will be at sta. 109 ; 
at 110 set a stake by offsetting 8.72 ft. The last chord is 60 long, 
and hence the offset 

]^F:=^ 4.37 X ^ + 4.37 X (j^J= 262 + 1.57 = 4.19 ft. 

Make EQ = 1.57 ft., and prolong QF, the terminal tangent. 



LOCATION. 



67 



H- 



95. To Locate a D Degree Curve by Offsets from Tangent. 

Let AMy Fig. 22, be tangent at A, and E, F, 0, etc., points on 
the curve. Tlie offsets BE, CF, ^ 
etc., maybe found from formula lT 
(86), 

z = J»«i), 

either by taking equal intervals, ^, 
AB, BO, CM along the tangent or 
by taking E, F, G, etc., at regular 
stations around the curve and 
using the arc length instead of 
the tangent. 

When the arc AO is large, or 
strict accuracy is required, we 
proceed to find the offsets at 
regular stations and the lengths 
of AB, AG, etc. First find E 
from (12) or (12'); then from triangle OEL, 




Fig. S2. 



BE= AL = R(l - COB D) = B vers D, 
AB= LE= B sin D. 
lu like manner 

CF = AH = R{1 - cos 2D) = R vers 2A 
AC^HF = B sin 22), 

and so on for any number of stations. 

Should^ fall at a plus station, we first find the angle i>i at the 
center, then 

BE-B vers Z>, , 

AB-B sin />, , 

CF - B vers (Di -f 2>), 

AC= 72 sin (Z>, -f /)), 

etc. = etc. 



The ordinates BE, CF, etc., are evidently equal to the mid- 
ordinates for long cl^ords 2LE, 2HF, etc.; hence we can. if 
A, E, F, and O, fall at full stations, take them direct from 
Tabic V; then take the long chords from Table IV and dividing 
these by 2, get the required coordinates. 



58 A field-majNUal for railroad engineers. 

ExAMPLB. — Locale three stalious of a 4° curve by offsets every 
50 ft. on curve. 

Referring to Tabic V, the required offsets are 0.87, 3.49, 7.85, 
13.94, 21.77, and 31.31. By Table IV the distances measured 
along tangent are 50.0, 99.94, 149.76, 199.39. 248.78, and 297.87. 
With these values we can set out the curve either way from A. 

Had we used formula (36) we should have had for the values 
of the offsets 0.87, 3.50, 7.88, 14.00, 21.87, and 31.50. 

96. To Locate a Curve by Ofibets from a given Liong 
Chord. 




Let FK, Fig. 23, be the given chord. We may compute the 
offsets yi , y« . . . Jtf by the methods of 83 — of which formula (17^ 

is the most convenient, within the limits of its applicability — 
and setting off these ordinates, locate the curve. 

Or w^ may setoff the mid-ordinate M= R^GvaFOA at A, 
and at C set off y, = Jf — i? vers D, making 



GEMWhe 



AC = HL = E sin D. 



M- R vers 27), and AE = R sin 2i>. 



Another Method is to find the angle KOFhi the center, and 
by Table IX deiermiue BA^ M\ then by Tables V and IV 



LOCATION. 69 

determine BL, 5.V. LH, and ^'G. Then tiC =^ M - BL, which 
sel off at C, uud other points in like manner. 

Example.— Given the P.O. of a 4** curve at station 160 + 75, 
the angle between tangent and chord = 9', i-equired the offsets 
necessary to locate the curve. 

Here 7=2x9 = 18^ 

18 
.-. X = T = 4.50 stations. 
4 

Hence the RT, falls at 160.75 + 4.50 - sta. 165 -f- 25: The 
mid-point on curve B falls at sta. 163. By Table IX, 

3,= !?^ = 17.64 ft. 
4 

By Table V the mid -ordinate for two stations of a 4"* curve is 

BL = 3.49. 

Hence HG = 17.64 - 3.49 = 14.15. 

By Table IV, HL = AC= 99.94 ft. 

Measure AG = 99.94 ft., and set off' GH— 14.15 ft., and drive a 
stake at E. In like manner find 

(?i?=3.70 and ^Z?= 199.3J ft. 

The points P and Q are also located by means of the coordi- 
Yiates just determined. 

If B had fallen at an odd station, the curve could have been 
located in the same manner, ^and P being 100 ft. from B^ Q and 
q 300, etc. 

97. To Locate a Curye with Transit and Chain when the 
Degree D or Radios R is Known. 

If R is given, determine D by (13); then, since the angle in 
the circumference of a circle is half the angle at the center sub- 
tended by the same chord, we may locate points- on the curve by 
successive deflections from the tangent. 

In Fig. 24 let the P. G. be at A, at which point set the transit, 
and with the vernier plates clumped at zero place the telescope 
in tangent either by sighting the P.L or by backsighting to some 
point in the tangent Deflect from the tangent half the angle at 
the center for the sub-chord or chord, and direct the head chain- 
man into line while the rear chainman holds his end of the chain 



60 A FIELD-MAKUAL POH RAILROAD ENGINEERS. 

at the trausit, the cbaiu being kept taut. The stakeman drives a 
stake at the point where the head cbaiuman's flag rested, and the 
rear chain man advances to this point. Deflect ^D from the chord 
AB just run, and while the rear chainman holds his end of the 
chain at B direct the head chainman into line at C. Other points 
are located by deflecting an additional ^D for each chord length 
measured, until a point E is reached to which it is desirable to 




Fio. 24. 



move the transit. The angle FAM should not exceed about 15". 
Move the transit to E, backsight to Ay and deflect FEA = EAF, 
when the telescope will be in tangent, and the curve can be con- 
tinued until it is again necessary to move the transit. At the 
P. T. put the telei^cope in tangent by backsighting to the point 
last occupied by transit and deflecting the tangential angle as at 
E, The line may now be continued. 

98. The Index-angle is read on the vernier-plate, and is the 
angle between the tangent to the curve at the P. 0. and any other 
line passing through a point on the curve when the telescope is 
directed along this line. It is most frequently taken as the angle 
between the initial and any subsequent tangent to the curve. 
Thus at E the index-angle equals EFP = 2FAE. At any point 
on the curve the index-reading in tangent may be found by the 
following rule, which may be easily deduced from a figure: 

From double the index- angle that fixed the point subtract Vie index- 
angle in tangent at the last point; tJie remainder is iJie index-angle 
required. 

99. Subdeflection-angles may be found by (13) rigidly, or 
approximately (and with sufficient accuracy except when D is very 
large) by assuming the central angles to be proportional to their 
chords. Thus on a 4*" curve the central angle for a sub-chord of 
25 ft. would be T, and the subdeflection angle 30'. 



LOCATION. 61 

Example -Locate a 4** curve to left when the P.C. is at sta. 

81-f-25and/=B2'86'. 

Here L ^ — f- = 8.15 chains. 

4 

Hence the P. T. will fail ai 81.25 + 8.15 = sta. 89 + 40. The 
first sub-chord is 75 ft. lon^, hnd the first deflection-angle will be 
found by (18). 

.-. ^5 = 1** 30'. 
By the approximate rule, since ^D = 2"*, 

2 ""100' 

whence i^ = 2 X f = 1** 30' as before. 

With transit at P,C. deflect 1* 30' from tangent, measure 75 
feet, and set sta. 82. Then a deflection of 3** 30' will determine 
83, 5** 30' sta. 84, V 30' sta. 85. Now remove transit to 85. and 
with vernier at T 30' backsight to 81 + 25. Reverse telescope 
and set vernier at 15° 00', when the telescope will be in tangent. 
An index angle of 17" will fix 86, and so on. 

The last chord will be only 40 feet long, for which the sub- 
deflection -angle is yVtf ^^ 2°> ^^^^ ^^' ^^'' '^^^ index-angle fixing 
the PT. is therefore 23" 48'. 

To get in tangent at 89 + 40 backsight to sta. 85, with vernier 
at 23° 48' ; then by the rule of 98 the index-reading is (23° 48) X 
2 — 15° = 32° 36' = /. Set the vernier at this reading and run 
tangent. 

Caution. —It is not good practice to set more than 4 or 5 sta- 
tions on curve from any' one point. Mr. Shunk gives the limit- 
ing angle to be deflected from tangent as 20°, and says 15° should 
rarely be exceeded. (Field Engineer, p. 82.) 

100. The Transit Notes may be conveniently kept in the form 
below, which shows the notes for the last example. 

When possible the tangents should be run to intersection, the 
angle I measured, and the tangent distance calculated. Then 



62 A FIELD-MANUAL FOR BAILROAD ENOINEERS 



Station. 


1" 


^1 




ii 


. 


Remarks. 


00 
















-1-40 


QP.T. 


0«48' 


23043/ 


82«86' 


N27«»88'E 


N27'»30'E 




89 






230 0' 










88 






21° (y 










87 






19« 0' 










86 






17«» 0' 










85 


O 




7° 30' 


15* 0' 








84 






5° 30' 










88 




2«> 0' 


8** 30' 










88 




1«30' 


1«30' 








4'» C.L.; P.I. set. 


+!» 


O PC. 4*C.L. 


0«» 0' 


O* 0' 


0» 0' 






J = Si' 36'; T ^ 
418.9 ft. 


81 










N(K)«»12'E 


NeO'lCE 





measure along tangents and set P.C. and P,T. from the P.I. 
Wbeu tbe curve is run in, the position of the P.T. thus found 
should agree with the one set from the P.I. If the error is 
greater than the circumstances of the case permit, the curve 
must be rerun and tangents remeasured. 

101. Another Form of Notes, and in some respects a better one 
than the above, is given below. The index-readings are com 
puted as though the entire curve were run from the P.C. The 
notes for ihe last example would appear as below ; 



Station. 


II 


-J 
II 


Si 

p 


h 
r 




Remarks. 


90 
















+40 


0P.5. 


0«>48' 


le-is' 


32«>36' 


N27'»36'E 


N27«»30'E 




89 






15<»30' 










N8 






13«30' 










87 






11«80' 










8() 






9«30' 










85 







7*>30' 










81 






5*>3C' 










bQ 




20 0' 


S'SO' 










8-J 




1»30' 


1*>30' 








!• curve left; 


-f25 


OP.C7.4«'C.L. 


0« 0' 


0" 0' 








PI set. i=32<»36'; 
r= 418.9 ft. 


81 










N 60*12' E 


N60»10'E 





The computations are all made before beginning the work, and 
the notes have the advantage of permitting the tracing of the 
curve either way from the instrument without additional rompu- 



LOCATION. 63 

tations. Suppose the traiisitmaD to have run the curve from the 
P.C. tosta. 85, lo which point he removes the instrument. He 
there set« the vernier at 0"— the ttngle on limb when telescope 
was in tangent at i he P.C. — then sighting theP. (7. he reverses 
the telescope and deflects to 9' 30', which will fix sta. 86. Had 
the tangent at 85 been desired, a reading of 7" SC— the angle that 
located that point — would have put the telescope in the plane de- 
sired. A reading of IT 30 fixes 87, and so on to the F.T. 
Removing to the P,T., the plates are clamped at T 30', and a 
backsight to sta. 85 taken ; then deflecting to 16** 18', the tele- 
scope is in tangent at the P. T. Had it been desirable to set 84 
from 85, a reading of 5° 30' would fix that point ; others may 
be found in the same manner. 

Any convenient form of notes, which are intelligible to another 
engineer who may have to retrace the curve, may be used, but it 
is desirable that some general form should be employed. Either 
of the preceding forms seems to meet ordinary requirements. 

C. Obstacles. 

102. To Pass an Obstacle on a Curve. 

FmsT. Suppose Uie obstacle to he one obstructing vision at one 
station only. 

In Fig. 25 suppose transit set at A, and B and C located from 
that point, but the next full station, 12, to be invisible from A, 



Fig. 25. 

Set a plus station at E, as near the obstruction as may be conven 
ient, then set FlOO feet from E. Next make FG= 100- CE, 
and locate O with the corresponding deflection-angle. Other 
stakes may be set beyond G, or the transit may be removed to 
that point and the curve beyond traced. 

Second. Suppose the line of sight qhsi^ured for more than me 
station, as in Fig. 26, 



64 A FIELD-MANUAL FOR RAILROAD ENGINEERS. 



If trausit is at Ay deflect an angle HAB that will clear all o\ 
structioDS, aud at the same time cause B to fall at a full statioa. 
Then by Table IV, Table IX, or by formula (16) calculate the 
long chord AB ; measure ^2? and move transit to B ; then defect 




Fig. 26, 
the angle ^BC=jR4-ff when the telescope will be in tangent. 
The curve may now be run both ways from B, 

If it happen that some stations, as E and F in the figure, are 
still invisible, they may be located by offsets from chord or tan- 
gent. 

Example.— Let the curve be a 3* curve to right ; angle HAB 
= 7** 30', the deflection-angle for 5 stations. By Table IV the 
long chord is 498.63 feet, which can now be measured and a hub 
set at B ; then making angle CBA = 7'' 30', the telescope will bo 
in tangent and the curve can be traced either way. 



103. To Locate a Curve when the P. G. is Inaccessible. 

^ In Fig. 27 let the P.C. at B be in- 

Q 




FiQ. 27. 



accessible ; it is desired to reach a 
— point H on accessible ground. 

First Method. — Assume a 
point jET on the curve such that a 
line AH from an accessible point 
A, on tangent, will clear the ob- 
stacle ; for convenience H should 
be at a full station. The arc BH 
aud central angle, which equals 
ECFy are then known. Calculate 
BC=Thj (14) or (14a) ; then 
since AB is known, AC, = AB-^ 
BCy is known. 

Now in triangle il (7^, from trig- 
onometry, 

tani(7i-a) ^ AG - C H 
tan 4(7* -fa) ~ AG+GEC 



LOCATIOK. 65 

But (h + a) = c; hence 

ten4(A-a) = ^J-p-^tanlc (40) 

Then i(A + a) + i(A — a) = h, the larger angle, and 
4(A -fa) — J(A — a) = a, the smaller angle. AH may be 
found by the law of sines, or by drawing CE perpendicular 
to AH, when 

AH = AG cos a + CH cos h (41) 

Example.— The P.G. of a 4° curve is at sta. 141 -f 25, and it 
is desired to reach the point ^from sla. 139 on tangent. 

Suppose H be assumed to fall at sta. 147 ; the curve length ic 
X = 147 - 141.25 = 5.75 chains. Then angle c = 5.75 X 4 = 
23" (y. By Table IX the tangent distance for a V curve is 
Tx i- 23'' = 1165.8 ft. 

By (14a), T = ^-^^ = 291.45 ft. 

Now AC= 291.45 + 225 = 516.45 ft., 

and 

AG+ CH= 516.45 + 291.45 = 807.90, 
while 

AC- (7fl^= 225ft.; 
hence, by (40), 



tan i{7i - a) = -^7r=-^ X 0.20345 = 0.05666 = tan 3" 15'. 
"07.9 



Therefore 
and 



A = 11** 80' + 3° 15' = 14** 45', 
a = 11'' 80' - 3'' 15' = 8' 16'. 



By (41), 

AH= 516.4£» X 0.98965 + 291.45 X 0.96705 = 793.0 ft. 

At A deflect 8" 15' from tangent, measure 793.0 ft. and set a 
hub ; move to this point, backsight to .1 and deflect 14" 45' into 
tangent, then trace in the curve. 



66 A FIELD-MANUAL FOR RAILROAD ENGINEERS. 

Second Method. — If F, any assumed point in tnngcnt, is 
visible from J., AF may be measured by some indirect method; 
then AF ^ AB = T, The tangent for a 1" curve having same 
iutersection-angle, KFG, is Ti = TxD; find this value of Tx in 
Table IX and take out the corresponding- value of 7. With 
transit at F deflect the angle KFG, measure FO = FB = T, and 
set hub at O, The station number of G is found by dividing the 
central angle, = KFG, by the degree of curve Z>. Move to G and 
trace the curve. 

Example.— Let AF measure 490.5 ft. from sta. 139 of the last 
example. Then^J5 = 325 ft., and BF= 490.5 - 225 := 265.5 ft. 
265.5 X 4 = 1062 ft., which by Table IX is the value of T, for 
/= 21^ Set transit at F, deflect 21% and measure FG = 265.5 ft. 



L= -7-=^ 5.25 chains; 
4 



hence G will fall at 141.25 + 5.25 = sta. 146 + 50. Move to Q 
and run the curve both ways. 

Third Method.— In Fig. 28 let the inaccessible P.C. be at B, 
and let it be required to reach E from a point G on the curve 
prolonged backwards from B. 

At a given point A on tangent cal- 
culate the tangent oifsel by (86) or 
the methods of 95, then set this off at 
right angles to AB ; set the transit at 
C and turn oS ACL = 90<» - COB, 
when the telescopy will be in tangent 
at C. COB may be found from Table 
IX by multiplying AC by the degree 
of curve and taking half the iutersec- 
tion-angle corresponding to the mid- 
ordinate that equals this product. Now deflect and measure 
ECL, then by (16) or (16«) calculate CE, which mejisure. Move 
to E and deflect LEG = ECL and the telescope will be in 
tangent. The central angle BOE = 2LEC - BOG, from which 
the arc 5^ and number of sta. E may be found. 

Example. — Take the same example ns in the last two cases. 
A is at sta. 139, B at 141 -I- 25; hence AB = 2X5 stations. 




Fio. 28. 



By (36). z = AC=lX{'^ 2CV' x 4 - 17.72 ft. 



LOCATIOK. 



67 



Or by Table IX the angle correspondiDg to the long chord 

(2 X 2.25) X 4 ^ 1800 ft. is 18'' 4', for which the mid-ordinale is 

71 06 
71.06 ft. For our 4" curve the mid-ordiuate will be —^ =: 17.77 

ft., whicb equals AG and agrees closely enough with the value 
for z above. 

Make angle BAG =^90\ and measure -4C = 17.72 ft. Move 
to G and sight to A, then make angle ACL = 90" — (9*2') = 
80'' 58'. Suppose an angle LCE = 16** 1' to clear the obstacle. 
By formula (16), 

GE = 2R sin (16*' 1') = 2 X 1432.7 X 0.27592 = 790.6 ft. 

Measure along GE 790.6 ft. and set a hub; move to E and run 
the curve. 

GE might have been found by means of Table IX, for the long 
chord of a 1' curve having i = 2LC^ = 32" 2' is 3162.0 ft.; 
divide this by 4 and there results GE= 790.5 ft. 

104. T9 Pass to Tangent when the P.T. is Inaccessible. 

This is just the reverse of the preceding problem, and may be 
accomplished by reversing the processes described above. 

When the P,T., however, falls in or beyond a river or lake 
obstructing the ordiunry methods of indirect measurement, the 
case merits a f^pecial solution. 

FiRsr MuTHOD —In Fig. 29 let the transit be at A, and B the 

P.T. From tiie known station 

numbers of -4 and B the length of 

curve and angle / may be found; 

then, by (14), Aa= R tan J/, or, 

T^'4-P 
hy{Ua\AC= -^-^. 

Move to Cand deflect the angle 
/; set a stake F, and one at some 
other accessible point E\ measure 
angle .ECF = c. Move to F and 
measure the angle EFG and the 
side EF\ then in triangle EGF 
angle e = 180" - (c +/); by trigo- 
nometry 

CF=^^.EF. (42, 

sm c ^ 




QS A FIELD-MANUAL FOR RAILROAD ENGINEERS. 

Siuce BC=zAC, there results BF= CF- ^C;'aDd as the sta- 
tion Dumber at B is known, that at F becomes known, and the line 
may be continued. 

If B is not the P,T., measure back the distance FB, set transit 
at B, and continue the curve. 

Example.— Let the P.T. of a 2° C.L. fall at sta. 205 + 50— an 
inaccessible point; suppose A at sta. 200, angle c = 40°,/= 80% 
£'2^=310 ft. 



Here 



i=5.50x2 = ir 0', and d = 60\ 



By (14a), T = ^5?^ =. 275.87 ft. 



From (42), applying logarithms. 



log (7F= 2.49136 -f 9.93753 - 9.80807 = 2.62082. 



Whence CF= 417.7 ft. Then BF^ 417.7 - 275.87 = 141.8 ft. ; 

therefore the number of i^ will be 206 4-91.8. 
Second Method.— In Fig. 30, with the transit at any point A 
on ihe curve, assume a long chord AB 
and calculate the angle CAB\ deflect 
this augle from the tangent AG^ and set 
a point ^ beyond obstruction ; set also 
a stake at C in tangent. 

Move to E and measure AEG and 
side EG. Compute AE from the trian- 
gle AEG. If this is greater or less 
than the kngth of the long chord AB, 
take their difference BE and set a hub 
at B. With the transit at B tmce out 
the curve. 

Example. — Given A at sta. 210 of a 
3"* C. L., angle a = 12^ b = 92% EC 
= 181 ft. Then c = 76% and by solving 

the triangle AEG. AE= 844.7 ft. By Table IX the long chord of 




Fig. 30. 



a V curve for /= 24*^ is 2382.6 ft. ; therefore AB = 



2382.6 



= 794.2 



ft. Now will ^5 = 844.7 — 794.2 = 50.5 ft., which is the dis- 
tance along EA that transit must be moved back from E, 



LOCATION. 



69 



106. Given the Perpendicular p from a Point to a Tangent, 
to Find the Point on Tangent at which to Begin a Curve of 
Given Radius which will Pass through the Given Point. 

First Solution.— In Fig. 81 let P be the point, BP the per- 
peudicular. We have to find 
BA=x, ^^ 

From P draw PC parallel to 
AB ; then in triangle OPG 



ip = «* + (5 - py. 

From which 



X = i^2Rp - pK 



(43) 



Second Solution.— Consider 
p = AC &s the mid -ordinate for 
a long chord = 2x ; then pX I> 
= the mid-ordinate for a 1" curve 
for a central angle equal 2a. 
The coiTesponding long chord may be taken from Table IX. 
Then 




Fio. 31. 



i x.q, 

2 D • 



{48a) 



ExAMPLE.-Given ;? = 80 ft., 2> = 4* (5 = 1433.7), to find a?. 

By (43), X = V85,962 - 900 = 291.65 feet. 
By the second method, 

30x4 = 120, 

the mid-ordinate for a 1* curve corresponding to an angle o' 
23'* 29', for which the long chord is 2332.6. Now. by (43a), 

fi 4 

106. In Fig. 31, Given x and p to Find the Radius of a 
. Curve Tangent to AB at A and Passing through P. 



From (43), 



i? = 



_ aj* + p* 



(44) 



70 A FIELD-MANUAL FOR KAILROAD EKGINEfiRS. 

107. Given the Location of a Point P referred to the P. L 
to Find the Radius of a Curve through P which will Unite 
the Given Tangents. 




Fio. 32. 

In Fig. 32 suppose BC — Z, BP = vi known, and angle a cal- 
culated ; or PC nnd a may be measured on Ihe field. 
From triangle CAO, 

6 = 90'' - (a + J/), and CO = i? sec JZ 

Now from triangle PCO, 

s»n y = ^ sin ft. 



Inserting values of PO and GO, 



sin y = 



___ R sec \I 



B 



. sill ft = sec J2 . sin 6 = 



sin ft 
cosfT' 



(45) 



: ?i eqiiation from which the unknown B has disappeared. Nexi, 
from the same triangle, since x = 180** — (ft + y), 



sm X 



"When / = 90*, it can easily be shown that 
B = l-{-m + V2im, . 



(46) 



(47) 



LOCATION. 



71 



108. To Locate a Tangent to a Ounre from an Outside 
Point. 
First Method. —In Fig. 83 let P be the point and AHB the 



E H 




rurye. Run a trial-line PA cutting the curve in A and B, 
Measure PA and AB ; or measure PA and angle a between the 
chord AB and tangent AL, Then 

AB = %AC = 2B sin a, 
OC = R cos a. 



By geometry, PE = \^PA X P^B, P£7 being the required tan- 
gent. From the figure, 

GO 
CP 

PE' 



tan » = 
tan m - 



At P deflect the angle I = m— n from PA and run the tangent. 
Second Method.— In Table IX find the long chord for t 
central angle 2a ; then 



AB:=z2AC = 



CH=^. 






and CO=:B- CH, 

We may now proceed as before. 



72 



A FIELD-HAXUAL FOR RAILROAD EKGIXEERS. 



1 



109. To Run a Tangent to Two Located Curves of Contrary 
Flexure. 

First Case.— In Fig. 34 let FK and LE be the curves, and 
KL = 'p measured on the ground. 




Fig. 34. 

Let FE = ^ be the required tangent. 

Draw OxH parallel and OiH perpendicular \XiFE\ from the 
triangle Oi ifOa , since FH = Hi , 



whence 



«= i/2(/?»-f /?,)!> +p». 



(48) 



Also. 



cos a = 






(49) 



The arcs FK and LE may be found from tbe angle a and the 
known curvatures, after which tbe points^ and Em&y be set. 

If ^ is given and p required, it may easily be found from (48). 

Second Case, p not kiunon. 

Set the transit at a point A on one curve and note the bearing 
of ibe laugent to the curve at that point (see Fig. 34); the bearing 
of the radius 0%A differs from this by 90**, Run a line ABC of 
one or more courses to intersect the other curve at C, Note the 
bearings and lengths of these courses and the bearing in tangent 
at C, from which calculate the luariiig of C<fx 7^, and R^ being 
known, the latitudes antl uepnriuics arc next calcuhitcd. Let OiJV 



LOCATION. 



73 



be the sum of the northiDgs or southings, Oii^Tthesumof the 
eastings or westings ; from the triangle Oi OtN, 



tan& = 



OtN' 



and 



OiO, = VOxN^ 4- 0^N\ 



As before, FE is the required tangent and OsS* perpendicular, 
while Oi^ is parallel thereto. 



cos a = 



Rx+Rt 



Angle FO^N= 6 - a is the bearing of O^F, while AOtF= 
e — b-\-a is the angle of retreat from the known jwint A to F, 
where the tangent may be run. The length of t= OiHis 

t = 0,0i sin a. 



D. Change of Location, 

110. To Locate a Curve Parallel to a Given Curve. 

Let p be the perpendicular between parallel tangents, and sup- 
pose ABC located (see Fig. 35). 
If there are no restrictions as to the 
position of the points By F, and G 
on the second curve, we may cal- 
culate the new degree of curve Di 
for a radius /?, = i? -f p, by (13), 
and trace the curve from any 
point, as E, Thus 



iA = 



50 
R^ 



50 




R + P' 

If, however, points on the radii 
through A» B, and G are wanted, 
they are gotten by using the same degree of curve D and com« 
puiiug the length of chord FE. From similar triangles, 



Fio. 85. 



EF 

Ri '' 



AB 

R '' 



100 
R' 



^ H-i A FIELD-MANUAL FOR KAILROAD EKGlNEfiRS. 
whence 



J?i7» = 100 ~ = 100 :?i^ = 100(1 4- 1). . . (50J 



Had EFO been the located curve, with radius R, we should 
have had 



^5 = 100 






(51) 



111. To Change the P.C. of a Iiocated Curve so that P.T. 
will Fall in a Given Tangent Parallel to Terminal Tangent of 

Located Curve. 

Let AB, Fig. 86, be the \o, 
cated curve; FE, the tangent 
ill which the P.T, rnubt fall. 

Let the distance between tau- 
geuts be HE = p. 

Draw BE and 00' parallel to 
AF; evidently AC = 00'^BE\ 
0' being the new position of 
center. 




In triangle BEE, 



BB-AC- -j?-^ = p cosec /. 
sin/ ^ 



(52) 



Set the new P. C. by measurement from A, and run the curve 
CE. A.ny system of straight lines and curves may be treated as 
above, provided 1 is the angle between initial and terminal 
tangents and p as before. 

Example.— A located 2*' 80' curve, having / = 25', ends in a 
tangent 25 ft. outside of desired tangent. Find the change in 
position of P.C. 



By (52), 



AC = 26x 2.86620 = 59.16 ft. 



LOCATION. 



75 



112. To Find the Change in Raditu and Position of P. (7. if 
P.T. ia Required to fall on the same Radial liine but on a 
Tangent distant p from, and parallel to, Terminal Tangent to 
Ijocated Curve. 

In Fig. 37 let AB be the located and CE the required curve. 
Draw the parallel chords AB and 

CE. Draw CA'and^-F perpendicular .A^. C_K 

to AB. The &ng\e8 FBE:=CAH-iI^ 

From the figure, 

CH= AG sin ^I, 
BF - BE cos \I = p cos \L 
Equating, p - 

-4(7 sin J/ = p cos J/, 
whence 




Fio. 37. 
4(7 = p cot J/. 

In the triangle OPOu dP = AC, OP = B - Bi, and 
(B—Bi) iSLuI=zAC = p cot i/, 
or B- Bi = AC cot I =zp cot J/, cot /. 

Therefore 

i?, = 12 - -4Ccot / = i2 - p cot J/, cot 7. . 

From trigonometry, 

^ , -. sin / , ^ , cos / 

cot i/= , and cot 7 = -; — =. 

1 — cos 7 sm I 



(68) 



(64) 



Insertiug these values in (54) gives 
sin 7 cos 7 



Bt=B--p. 



= B''P 



cos 7 



1 — cos 7 ' sin 7 
From trigonometry, ex sec 7 = 
• '• Bi = 7J — 



1 — cos I 
vera 7 



'B-p- 



cos 7 



vers 7* 



cos 7' 
P 



e* sec /' 



(64) 



76 A FIELD-MANUAL FOR RAILROAD ENGINEERS. 



Example. — A 2° SO' curve strikes 25 ft. inside a tiingent in 
which the P. T. must faU. Find the necessary change in radius 
and position of P.C. when / = 26^, 

By (53) the change in P. G. is 



By (54'). 



AC-26X 4.51071 = 112.77 ft. 
25 



i?, = 2292.01 



.10338 



= 2050.38 ft. 



By Table I we find this to be the radius of a 2* 47' 41" curve. 

113. Given a Located Curve uniting Two Tangents to 
Find the Change in Position of P.O. or in Radius for a Given 
Change in the Intersection-angle. 
First Case. — Radius unchanged. 

In Fig. 38 let BCE = / be the origi- 
nal intersection -angle, FCE = /' the 
new angle. From ihe figure, 
AO =^ AG- OG, 



AG = Ii (tan J/ - tan J7'). (55) 

By Table IX.— From the table, foi 
angle /, 




T = 



Fig. 38. 



For/' 



r = 



Then 



D * 

AG= r- T'. 



Second Case.— P.O. unchanged. 

Here the tangent T for the two curves is the same, and 
therefore 

iJ, tanii' = /?tanli; 

. (56) 



Whence 



ijj = /?tt.ni/.cot jr. 



LOCATION. 77 



By Tabub IX, 



whence D. = 'ElAll . jy = ^» ^ \ 

1 14. To Find the Change in i? and P. (7. for a Given Change 
in /, the P.T. remaining unchanged 




/O, 
Fig. 39.. 

In Fig. 39, from the triangles OBQ and OxBH, 
00 =z R cos / 
and OiH= Ri cos /i. 

Now QA = HF\ hence 

i?i - i?i cos /, = i2 — i? cos L 
Whence 

i?.=5[z:i£i4=fil?!if (67) 

1 — cos i, vers /i • v / 

Also, FA^BQ^BR- BQ. 

Inserting values of BE and BOy there results 

2?*^ = i?, sin /i - i? sin /. (58) 

116. Qiven a liOcated Curve to Find the Change in R for 
a Given Change in 7\ I remaining unchanged. 

In Fig. 40, from the triangles OAC and OiEC» since 
EA = EC- AG, 



78 A FIELD-MANUAL FOR RAILROAD ENGINEERS. 



Ml tan il-- Ri&u 11= EA = T' - T, 

Whence jB. = i?+ (T' - T) cot J/. (59) 

_E A c 




Fig. 40. 

By Table IX,^EA being known, T' = T + EA. Then, by 

(15a), 

__ T,^r 

"" — T' — ' 

If the change in vertex of curve is wanted, there results, from 
(25), 

^ = (7(? = rtau i/, E' - CH= T' ton \L 
Therefore GB = E' ^ E = (T' ^ T) tan iL . . . (60) 

GHcau be found from Table IX after finding Di as above. 
If Ri is given and EA wanted, (59) yields 

EA = T' - T = (Rt - R) tan J/- 

116. To Find the Radius of a Curve having the Same P,C, 
as a Given Curve, but ending in 
a Parallel Tangent. 

In Pig. 41 let the perpendicular 
distance between tiingents be p, and 
AB be the located curve; AOi = Ri 
is required. ; 

First Method. — Draw OH at 
right angles to OiE; then 

OiE= OJI+HQ+QE, 
or 

R, = (R, - i?) cos /+ i? + p. 




Fig. 41. 
From which 7?, = 7? + 



1 — cos 1 vers / 



(61) 



LOCATION. I'd 

Second Method.—^, B, and E lie on the same straight line, 
since / is the same for both curves. In triangle BQE angle 
EBQ = J7, and 



^^=8l^ = ^"^'^*^- 



From Table IX, AB = h^i^lJl, 



AE = AB 4- BE is Ibe long chord for curve of degree Dx\ 
therefore 



'" AE ' 



If desired, R may be found by (12') or Table I. 
Third Method. — Draw FL parallel to OiE\ then 



CF = -——. = » cosec 7. 
sini 



From Table IX. AG^ —ff^' 



AF=i AC-i- CF, the tangent distance for second curve ; hence 



2>,=- 



AF 



Remark. — If transit is set up at -B, it will be well to set E 
by measurement from B, to serve as a check when the curve is 
run in from A. 



80 A FIELD-MANUAL FOR RAILRDAD ENGINEERS. 



Article 9. Compound Curves. 
>l. Location Problems. 

117. QiTen Two Unequal Tangents, their Intersection-angle, 
and One Radius, to Find the Other Radius of a Compound 
Curve uniting Tangents. 

In Fig. 42, AH= Ti and Bn= T^ are the known tangents, 
AOi = lit the known radius. BOn = R^ and the angles /i and /» 
must be found before curve can be located. 




Fio. 4:4. 

Extend first branch to F, so that tangent FL is parallel to BH. 

Draw UK Aud BG perpendicular to FL ; draw FB and extend 
to ^; it will pass through the P.C.C., because the central angles 
EG, Fund EO^B are equal Then 

To = ^X = ^1 tan \L 

In triangle LHK, since LH = To— Ti, 

8 = KL^iTo- Ti)cosl 

p = HK=BG = (To~ r,)sin/. 

Now in triangle BGF Angle BFG = i/j, and 

l^FG=T,'[-S''T^, 



LOCATION. 81 



1 



taiii/, = ^. 



Draw OsJf parallel to FL ; then 

(J2i-i?,)8mJ, = t 
whence 

H^^Rx^ -T-^ = ijj - ^cosec /a. . . (68) 
sin /, 

Had B% been required, the equation would have been 

i?i =i?, + iooflec/9. 

Evidently, L^I-* /,. 

In the field the points E and B may be located by running in 
the curve from A as stailing-point, or run the chord 

from A, and at i?^ deflect angle AFB = J/— J/, = J/, , measure 
FB-l9ec i/« And BE = 2Rt sin i/,. 

Example.— A 2* curve has the RC, at sta. 110, Ti = 590 ft., 
7\ = 511 8 ft., 7 = 30** 50'. Locate the curve. 

By Table IX, r« = 1580/2 = 7d0 ft. 

By formulas above, 

« = 200 X 0.85866 = 171.78 ft., 
i? = 200 X 0.51254 = 102.51 ft., 
i = 790 + 171.73 - 511.8 = 449.93, 

tan J7, = ^^ = 0.22784 = tan 12" 50'. 

Then /» = 30** 50' - 25' 40' = 5* 10'. 

44 ft 07 
E, = 2864.93 - ^f^' = 1833 feet. 



82 A FIILD-HAKUAL FOR RAILBOAD EKQINSERS. 

By Table 1 this is seen to be the radius of a 8* 7i' curve. 

The length of first branch is 258.3 feet, and of the second 821.3 
feet; hence the P.C.C. falls at 112 + 58.3, while the P.T. is at 
sta. 120 4- 79.6. 

118. aWen the Long Chord from P.C. to RT. of a Com- 
pound Curve, the Angles it makes with the Tangents and 
One Radius, to Find the Other Radius and the Central Angles. 

In Fig. 42 AB is known, as also the angles HAB — a and 
RBA = h. Two angles and one side of the triangle HAB are 
known, and the sides HA — Tx and HB = T^ may be found, 
after which the solution is the same as in the last problem. 

A solution may be reached in a different manner. I = a + h^ 
EAF = iJ = \{a 4- h), and BAF =z\{a + h) - a = J(6 - a). 
AF = 2i?i sin \I. In triangle BAF two sides and the included 
angle are now known, so BF an* angle BFA may be found; 
OFB =^ il^ = il " BFA. 

Then JKF = 2i?, sin J7a , 

and EB= EF - BF becomes known. 

Then EB = 2i2a sin J/, = 2R, sin J/, - BF, 

BF 



1 



whence 2?s 

Evidently Ix = I - h 



Rx 



2 sin i/a • 



(64) 



119. Given the Radii and Central Angles of a Compound 
Curve to Find the Tangent Lengths, the Long Chord from 
P.C. to P.T,, and the Angles it makes with Tangents. 

In Fig. 43 draw AE and BE from the 
P.C. and P.T. to the P.C.C., then 
calculate AE and BE by (16) or by 
Table IX. In triangle AEB angle 
AEB = 180 - {{h + /«). Two sides 
and the included angle being known, 
the triangle AEB may be solved for 
AB and the angles ABE and BAE\ 
then 

BAF=^BAE+\Tx, 

Fio.43. ABF:r=ABE+ i/.. 

The angle AFB of triangle ABF now becomes known and, as 




L0CATI02S:. 83 

AB is known, the sides AF = I\ and BF = jTj may be com- 
puted. 

120. Given the Iiong Chord from F.C. to P. T. of a Com- 
pound Curve and the Angles it makes with Tangents to 
Find the Radii when the Common Tangent is Parallel to Long 
Chord. 

In Fig. 43 let QHhQ parallel to AB, and OAB = a, HBA = h 
known. Then 

BAE = BAG = QEA = Ja. 
and ABS = EBH = HEB = J6, 

Also, AEB = 180" - l(a + ft). 

In triangle AEB, remembering that 

sin [180 - i(a + b)] = dn K« + *), 
ABsinjb 



and 



4^= . 

sin i{a -f 6)' 



^^^^^sin^a 



sin i(a + by 

Since AOiE = a and i^Os^ = ft, the radii Bi and i?s may be 
found from formula (16), or (16a). 

By (16), ^»-3-n,-p~27i^|«.3inj(«^jy . . . m) 

' "" sin lb" 2 sin Jft . sin \{a + b)' ' ' ' ^^' 

Example. —Required Ri and R^ , or Di and 2>t , when AB = 
900 feet, a = 12% ft = 15^ 

By (65), i?i =2407.0 ft. 

By (66), i?, = 1543.7 ft. 

FTom Table I, i>. = 2** 22' 50" and 2), = 3' 42' 44". 



84 A FIELD-MANUAL FOR RAILROAD ENGINBBES. 



B. Obstacles. 

121. To Locate a Point on aie Second Branch of a Oom- 
pound Curve when the P.C.G. is Inaccessible. 

Ordinarilj the second branch is located by setting transit at the 
P. C, C and ruuniug the curve from that point. An obstacle on 
either curve may then be passed by the methods given for simple 
curves. 

When the RC.C. is iuaccessible, 
,Ap locate the first branch from IheP.C 

and the second branch from the 
P. T,, if this latter point is known. 
When this is not the case proceed 
by one of the following methods: 
First. By meana of a long 
chord. 

In Fig. 44 let ^ be the P.C.a. 
A some known point on first 
branch, EF a tangent at B, and 
AB parallel to FE. The station 
numbers of A and B being 
known, the arc AB and angle a 
are readily found ; then 

F!L = /?9 vers ft = /?i vers a. 




whence 



next. 



vers ft = 



i?i vers a 



(67) 



AB = Rx sin a -f- 2?, sin ft. 



EL = 



(68) 

Defiect FAB = a from tangent at A i measure out AB ; set the 
tnuisit at B and locate the second branch. 

By Table IX. —Take the mid-ordinate in table for an inter- 
section-angle 2a ; then 

A • 

Then EL x A is the mid-ordinate for a 1* curve having 
/ = 2ft, from which ft becomes known. From the table now find 
AL and LB, the half-chords for angles 2a and 2ft, and proceed as 
before. 

Second Method.— £;V ^n^n* of tangents, 
JVom Fig. 44, AF= FE = ff. tan Ja. 



LOCATION. 



85 



Set transit at F, deflect OFE = a, and by some indirect method 
measure to an accessible point H. 



and 



BH=:FH^FE, 



EH 



tan \h = -5-, from formula (14). 



Angle h is now known and equals OHE, which deflect from 
BH\ then measure HB = EH, and with transit at B locate the 
second branch of curve. 

Or by Table IX.— Find AF = FE, the Ungent distance for 
I =z a; then having EH measured, take Ti = EH X !>% and find 
the corresponding angle, which equals b ; then proceed to locate 
curve as above. 

Example.— Let ^4 be at sta. 126, P. C. C. at 128 + 25; the degree 
of first branch 4**, and of second 6**. 

By the first method EL = 17.635 for a = 9^ and b = 11** 2'. 
nearly. AL = 224.1 ft., BL = 182.75 ft., and therefore AB = 
406.85 ft. Angle 6 = ir 2' corresponds to 183.9 ft. around 6- 
curve; hence the P.T. number is 130 + 08.9. 

By the second method AF = 112.74 ft. Suppose FH = 264 ft., 
then EH= 151.26 ft., which multiplied by 6 gives 907.56 ft, 
corresponding to / = 18**. The arc EB is now 300 ft., making 
^ fall at sta. 131 + 25. 



C. Change of Location. 

122. Having a Simple Curve Located to Find the P. C.G. so 
that a Curve of Given Radius shall connect with a Given 
Tangent Parallel to Tangent to 
Located Curve. 

Let NAB, Fig. 45, ,be the located 
curve, HF the tangent in which the 
second branch must end. The dis- 
tance BO = p between tangents is 
kno\yn from measurement. If angle 
a can be found, the arc BA becomes 
known and the point A can be located 
from B. Draw O^L from the center 
of second branch perpendicular to 
Pio.i5. OxB. In triangle OiO^L, OiOi = 

Bi - -Rj, and OiL = 5i - (1?, + p) ; therefore 



0% 


f ^ \ 


^ 


•^ 


L 




Rt 

F 


*-ip-^ 



86 A FIELD-MANUAL FOR RAILROAD ENGINEERS. 



cosa = 



Ri-Rt 



1 - 



Ri - R,' 



(69) 



Then a divided by Di gives arc BA. 

If desired, Bff may be found from the right triangle BEG, in 
which tlie side BO = p and angle OHB = Ja are known — 
A, ff, and B lying in the same straight line ; then 



BH=z 



P 



= p cosec |a. 



(70) 



sin ^ 

Or BA and 5-4 may be found from Table IX, after which 
Bff=BA-^HA. 

Example.— A 3*" curve ends in a tangent at sta. 160 + 50, 
35 ft. outside of desired tangent. Find the point of compound- 
ing with a 4' 50' curve. 

From Table I, R for 8** cui-ve equals 1910.08 ft., and for 
4" 50' curve 1185.78 ft. 

35 
Then, by (69), cos a = 1 - ;^ = 0.95168. 

From table of cosines angle a is found to be 17° 58'. Dividing 
this by 8 gives 5.961 stations for the arc BA. Hence the P,C,C. 
number is 160.50 - 5.961 = sta.154 + 53.9, and the new P.Jl is 
at sta. 158 + 23.9. 

123. Given a I«ooated Oomponnd Curve ending in a 
Tangent Parallel to, and a Oiven Distance from, a Tangent 
in which the Curve is required to end. To Find the Necea- 
sary Change in P. C C. 
First Case,— Terminal branch having shorter raditis. 

In Fig. 46 let ABC be the located 
curve. AEF the one required ; angle 
BOiC= a linown, and also MN = p. 

If angle EOM = b can be found, the 
angle of retreat from ^ to ^ will equal 
6 -a. 

Draw Oi'K and OiL perpendicular 
to ON, which is parallel to OiC, 




Fia.40, 



Then OK=(R^ R,) cos 6, 



LOCATIOK. 87^ 

Now £J/'= A, - -Si = iJ, - MN. from which KL=:MN^ p. 
Hence 

ifi — Ri) COB h^(R — Ri) cos a - p. 
From which 

cos ft = cos a — p ^ p (71) 

xl — /f I 

Divide 6 — a by 2>, the curvature of first branch, and move 
back that number of stations from B to the new P. G,G. at E, 

Join 0,0y'\ evidently FG = 0,0i\ aud angle KOt'Oi = CF'G' ; 
00i*0i = 90** - i(b - a), OOi'iT = 90** - h. Hence 

CFG = KO.'O, = [90 - 4(6 - a)] - (90 - &) = 1(6 + a). (72) 

From triangle GQF^ 

Or, from triangle OOx'Ox, * 

FG = Oi'O, = 2(5 - Rx) sin K* - «)• 

Had ^i^Fbeen the original curve, b would have been Icnown 
and a required. 

From (71), cos a = cos 6 + _ ^-, (74) 

H — JKi 

CFand angle GFMtire given by formulas (78) and (72). 

Example. — A 2" curve compounds with a 4" curve at sta. 
82 4- 30; a = 20^ 30', p = 40 feet. Find number of new P.G.G. 
aud distance between P.T.s, 

From (71), cos b = 0.93667 - ^^^ ^1432. 7 = ^'^S^^' 

This yields b = 24° 40', and 5 - a = 4° 10'. 
The change in P.G.G. is ^-r— = 2.083 stations; the P.G.G. 
number is therefore 82.30 - 2.083 = sta. 80 + 21.7. 



88 A field-makual fou railroad engineers. 



By (72), CFG = J(24' 40' + 20' 30') = IT 35'. 

By (73), 2?^ = 40 X 2.60399 =104.3 feet. 

Second Cabs. — The terminal branch having longer radiva. 
Let CAB, Fig. 47, be the located 
curve with FX\C. at A, and let 
FK be the tangent in which the 
curve is required to end. 

The distance BK = p, the radii 
OA = li, OxA = Ri , and angle 
AOiB = a being known, it will 
be sufficient to find angle EOx'F 
in order to get the angle of ad- 
vance, AOE -=1 a — h. Draw OL 
and OxN perpendicular to 0/F 
and OxB, From the triangles 




Fi«. 47. 
Oi'Oif and OxOL, 



But 



(2?i - 2?) cos h = Oi'iV+ (i?i - E) cos a. 
OiN—KB — p\ therefore 

(iJi -2?)co86=^ + (i?i — i?)co8a. 



Whence 



cos 6 = cos a + 



ifi-^' 



6( — * ( 

Then —j- will be length of curve from A to E. 



(75) 



Angle ^i^B = N0,0,' = 00,0/ - i^OiO. 
But OOxOx = 90" - \{a - 5) and i\rOiO = 90 - a. 

•. KFB = PO'' - \{a - 5)] - [90 - a] = J(a + ft). 
From triangle KFB, 



FB = 



= p . cosec i(a + 6). 



sin i(a + 6) 

Or, from triangle dOOi', since 0,0/ = FB, 
-TO = 2(/?, - i?) sin i(a - d). 



(76) 



LOCATION. 



89 



If AEF hekd been the located curve, b would have been given 
rind a required. From formula (75), 



cos a = cos b — 



Hi - B' 



(77) 



Example.— A 5** curve compounds at sta. 60 with a 2** curve, 
and the P.T. is at sta. 80. What will be the number of F.C.C. 
if the P,T. fall in a tangent 81 feet inside of terminal tangent? 
Here a = 40'. 



By (75), cos b = 0.76604 + -~^ = 0.81316. 



Hence b = 85° 36' and a - 6 = 4' 24', corresponding to 88 feet 
around the 6' curve. The number of the new P. G. (7. is therefore 
60 -f 88. 

angle KFB = 4(40" C + 36' 86') = 37' 48', 



and 



FB = Six 1.68157 = 182.16 feet. 



124. Given a Iiocated Oompoand Curve to Find Necessary 
Change in P.C.C. and Radius of Second Branch to make the 
P.T. fall in a Tangent Parallel to First Terminal Tangent 
and in a Point on the Same Radial Line. 

First Ca.sk,— Second branch Tiamng shorter radius. 

In Fig. 48, OB=B. OiB=Bi angle 
a aud HO = p are known. OiE=B% 
and angle b must be found ; then 

?^ = BE will be the change in 

p.ac. 

Produce first branch to K, where 
OiTis parallel to OiC. Since BOK 
= BOi C, B, K, aud C lie in the same 
straight line; and since EO%F ^ 
EOKy E, F, and K lie in the same 
straight line. Therefore 

KOG=:la, and KFH^\b. 




FzG. 48. 



90 A FIELD-MAKUAL FOR RAILROAD EKOIKEERS. 
Prom triaugles KFH &Bd KCO, 

, .. HK OK OH ^ . , p 



But 



FR=i OiL = (R-Ri) Bin a. 



. •. tao \h = lau \a + 



(2?--B,}8ina* ' 
Prom triangles OOxL and OO^M, 

(R - i?a) sin ft = (1? - /?,) sin a. 
^ sin a 



(78) 



"Whence 



i?, = i? - (2? - i?,) 



sin b' 



(79, 



Had ^i^F been the first curve located, h and J?8 would be 
kuown, a aud i?i required. 
From the figure, reasoning as before, 



and 



^°^ = ^°^^- (i{-^.)siut ' 



^ sm a 



(80) 



(81) 



Second CA8B.--^Sigc<?;i(i branch Mving longer radius. 




Fio. 49. 



In Fig. 49 let AB be the located curve, i^Fthe curve required 
OA = R, 0,A =. i?», O^S = i?„ FB = p. 



LOCATION. 91 



:^ 



f?a and angle b are wanted, angle a being known. 
We can show, as in first case, that 

OM=i KF= LB = (2?i - i?) sin a; 
and hence 

^ ^. HK EL p 

Or inserting values, 



P 

tan lb — tan Ja ~ -n ^r"* — • 

• . ■ (iii — 22) sin a 



d — b 
Angle b now becomes known and — jr— = AE in chains, which 

is the change in position of P. C. C. 
From triangles 00,Jlf and OO^M, 

(Ri - B) sin b = (i2, - i?) sin a 



.'.B, = B + (B.--Bf^^ (83) 



Had the new tangent fallen outside the old one, we should have 
bad 



tanja = tan46+^-^-^^^^, . . . (84) 



and 



A = ;?+(«. -fi)«^*. (85) 

126. Having^ a Located Oompound OtOrve, to Find the 
Change in F.G.C. and Radius of Second Branch in order to 
Cause P.T. to Fall at a New Point in Terminal Tangent. 

First Case.— Second branch having shorter radiue. 



92 A FIELD-MANUAL FOR RAILROAD ENGINEERS. 

In Fig. 50 let NAB be the located curve, and C the point where 
P.T, is required to fall. Let BG — k, 0A=^ i?, OiB = Rx, and 
angle Ox OH = a be known; angle b and R% are required. 




Extend first branch to Fy making OF parallel to OxB. A, B, 
and F lie on a straight line, for angles AOxB and ^OJ^are equal; 
likewise E, G, and F lie on the same straight line. 

From triangles OBF &nd OGF, 



^^^ OB GB ^, k 

coti6=^-.^=cotia-^ 



But OF = HM = (i? - -Bi)(l - cos a) = (R- i?,) vers a. 

k 



cot \h = cot Ja — 



(i^— iJi) vers a* 
From triangles OOi^and OO^L, since OiP = k, 
(B - /?,) sin b = (B-Rx) sin a - *. 
Whence 



(86) 



R^ = R- 



k-jR - Rx) sin a 
sin 5 



(87) 



Then b— a divided by D gives arc AM With radius i?, locate 
the curve i?(7 from C or ^. 



LOCATION. 



93 



Had NEC been the located curve, R, R% , and b would have 
been known, Bi and a required. In this case 



cot ia = cot Jd + 



/?, = 2?- 



{B— Ri)Yerab' 
A;-f(i?-g,) sin ft 



(88) 



Sbcond Casb. — Terminal brancli having longer radius 

In Fig. 51 let NAB be the located and NSC the required curve. 




Let CB = A; be known. Then, as in the first case, 



,,^ OC OB k 



.*. cot i6 = cot Ja — ' 



(i^,--K) versa' 
and (J?, - i?) sin a = (/?,- 2?) sin 6 + A: ; 



whence 



J? - » -L (-g»-^)8ina-A ; 



m 



. (91) 



di A FIELD-MAKUAL tOR RAILROAD ENGINEERS. 

Had NEC been located and NAB required, the equations would 
have been 



and 



cot Ja = cot \b -f 



Bi=zB + 



'-^(H. 


k 
- B)Yerab* 


(i?.- 


22) sin 6 + A; 



sin a 



. (92) 



. (93) 



In either of these two cases if k is unknown and the new radius 
given or assumed, the desired angle and the value of k may be 
fouud from the foregouig equations. Or, knowing the new angle, 
the new radius and value of k may be found from the sjiiiie 
equations. 

126. To Replace a Ourve of Oiven Radius, which nnitefi 
Two Tangents with Known Intersection-angle, by a Three* 
centered Compound Curve. 

In Fig. 52 let OA = /? be the radius of located curve 




Oa(7= OiA = Bi the radius of terminal portions of the three- 
ceiiiered curve, and the other notation as shown in the figure. 

Braw OaOa', and draw ^OiSTpei-peudicular thereto. From tri- 
ingles O^OiB and O^OH, 

O^H = (i?. — -Ri) sin J/i = (ij, - B) sin \L . . . (a) 

Suppose Bt and Bi to be assumed ; then equation (a) yields 



''^*^' = fclr«^»*^- 



(94) 



LOCATION. 96 

Then AO^' E = CO^O = \{I - I,). . . . (95) 

Suppose AO^'E, CO^Oy and R^ to have been assumed. From 
(95) find /i ; then, from equation (a), 

-B» = i?a - (iia - i?)-?^ (96) 

sm J/i ' 

Example.— Given a 4** curve, / = 38*, and the terminal 
brunches composed of a 2* curve for two stations, to find Bi and 
Dx for the central portion. - 

Here /i = 38° - 2(2 X 2)' = SO'. 

From Table I, i?a = 2865 ft., iJ = 1432.7 ft. 

Whence i?a — -R = 1432.3 ft. 

Log 1432.3 = 3.15603 
•* sin 19" 0' = 9.51264 



2.66867 
" sin 15** C = 9.41300 



.-. log 1801.7 = 3.25567 

Therefore Bx = 2865 - 1801.7 = 1063.3 ft., and, by Table I, 
Bi = 5" 23'. 4, nearly enough. 

127. To Substitute a Curve of Given Radius for a Tangent 
uniting Two Curves. 

In Fig. 53 let the tangent 5C=<, OBz=B, O^C = B^, and 
OtA = i?a be known. 

Angles a, b, and e must be found in order to substitute curve 
AE for the system ABCE, 

Draw OF parallel to BG, then O^F =B^'- B, and, from triangle 
00,F, 

tand = ^ ^^ , ^^'^ 

t 



00, = -T— T = «.cosecd= 4/(i?» --B)« + t». . (98) 
sin a 



96 A FIELD-MANUAL FOR RAILROAD ENGINEERS. 

Now iu triangle OOiO^ three sides are known and the angles 
c and e maj be computed. Thus if « is the half -sum of the sides, 



^ r 00, X o,o» 




Angle e may be found in like manner, then b = 180'— (« + <0» 
and a = c — 6. 

Points A and E may now be located and the curve traced. 

Example.— A 3° and a 5° curve are united by a tangent 60O 
feet long. Keplace by a 2** curve. 

Here R^-R= 1910 - 1146 = 764 feet. 

By (97). tan d = ^ = 0.65444 = tan 33** 12' 

By (98), 00, = 913.1 feet. 

In triangle OOiOa, 00, = 913.1, 0x0^ = 954.9, and 00, = 
1718.7 feet. Solving for e and c, 

tf = 183'36', c = 23'0'. Then 6 = 18" 12', a = 9" 48. 

Article 10. Track Problems. 

128. Reversed Curves should never be employed on main 
lines because of the shock due to sudden reversal of curvature 
and superelevation of outside rail. A short tangent should be 
interposed between the two curves, which may ordinarily be 
done by changing the end points of the curve, or slightly altering 
tb§ radius. If, Jipiypvgr, tn^nslUpn curves are employed to ease 



LOCATION. 



97 



off both curves, there would sceiu to be no objection to the use of 
curves of contrary flexure, provided the track may be kept 
always in perfect condition. In yards, crossovers, and where 
connection is made with existing track, reversed curves may be 
employed, and are often imperative. 

129. Having a liocated Curve Intersected by a Straight 
Une, to Connect them by Another Curve. 

Either the radius of the joining curve may be given, or else the 
point on first curve at which the junction must be made. The 
angle between a tangent to located curve at the point of meeting 
and the straight line must be measured. Four possible cases 
occur. 

FiKST Case. — Joining eurve tangent to located curve internally 
and on same Me of cvtting line ag center. 

In Fig. 64 let GF be joining curve, with center Oi and radius 
Bi, Let radius of located curve OF = B. Draw OiO and OH 
perpendicular to the cutting line produced, and OiK parallel to 
AH, If Bi is known, we must determine angle b, a having been 



R 
O5 


*^ i Q \a B 


♦'■d-y 


/ "a 


4 





Fio. 64. 
measured; then b — a gives the length of arc from A to -F where 
the P.CG. is to be located. In the triangle EOOi we have 

0K= OH- Bi and OOi = B - i?i. 

U cos a — Bi 



Then 



cos 6 = 



D 



B-Bt 
= arc AF. 



(99) 



Had Fbeen given, we should have b = a-\-AOF, and, from (99), 



iJi = 



B (cos a — cos b) __B (cos a — cos b) 



1 — cos b 



vers b 



(100) 



98 A FIELD-MANUAL FOR RAILROAD ENGINEERS. 



Example. — A l"* curve is cut by a tangeut that makes au angle 
of 64' 32' with tangent to curve. Unite by means of a 4* curve. 

By (99), cos 6 = 0.24000 = cos 76* 07', and therefore 6 - a = 
!!• 35', making AF, of figure, 11 58 stations. 

Second Case. — Joining curve tangent internally to located 
curve hut on opposite side of cutting line from center of located 
curve. 

In Fig. 54 let arc ME, with center Oi and radius i?a , be the 
joining curve. From the figure, 



coscf = 



R cos a + 2?« 



(101) 



Then arc ^^ = a - li divided by D, and c = ISO** - d. 
Had the point j^ been given and i^a required, it would have 
been, from (101) 

i?(co8 d — cos a) 



iJa = 



(108) 



1 + cos d 

Example.— Take the same example as in first case- Ilere, 

By (101), coBd = 0.9068 = cos 24* 56'. 

Then 64' 32' - 24' 56' = 39' 36', 

equivalent to 39.600 stations around curve from A to E, 

Third Case,— Joining curve tangent externally to located curve, 
toUh center on same side of cutting Une. 
Lf v*0. 




Fio.M. 



In Fig. 65 let arc 5(7,with crnter Oi and radius 7?, ,. be the join- 
ing curve. Draw OxE parallel lu CF, and OiC'aml OF pcrpeu- 
diculur thereto. 



LOCATIOK. 99 

From the figure, 

(R + Bi)co8b = Rcosa - A; 

.-. COS 5= B+Bi ^ ^ 

Then d = 180 — 5, and AOB = b — a. The curve may now be 

traced on the ground. 

If -4 C is wanted, we have AC = (/? -f 7?,) sin 5 — B sin a. 
If the point B is fixed and Bi required, there results, from (103), 

^^^ g(C0S«-C08ft) ^^^^ 

1 -|- cos 

Example.— Take the example given for the first and second 
cases. 
By (103), 

. 5730x0.43-1432.5 .^.. _,..., 

^^«' = 5730+1432.5 = ^'^^ = «^ ^^ ^^ 

J - a = 8r 44' - 64** 32' = IT 12', equivalent to 17.2 
stations on located curve from A to B. Angle d = 180** — 81" 44' 
= 98* 16', equivalent to 24.567 stations from 5 to C on the 
4** curve. 

FouHTH Case. —Joining curve tangent externally to located curve, 
with center on opposite side of cutting line. 

Let Oa, Fig 55, be center of joining curve, Bt its radius. 
From the figure, 

{B + 7f«) cos (J = i? cos « + Tfa. 

B cos a 4- Bi ,.^^, 

■--'^"'= B+H, ■ <^«^> 

If if is fixed and if, required, (105) yields 

P __ i?(cos c — cos a ) _ B(co8 c — cos a) . 

' "" 1 — cos c ~~ versin c * ' 

ExAMFLB. — Take same example as in preceding cases. 

By (105), cos c = 0.54403 = cos 57' 02'. 

Then « - c = 64" 32' - 57" 2' = 7" 30', 




r 



100 A FIELD-MANUAL FOR RAILROAD ENGINEERS. 

calling for a distance of 7.50 stations from ^ to Jf around 
1° curve. From M to Hon 4** curve is 14.258 stations. 

130. ToIiOcateaT 

A Y is made up of a system of tracks so arranged as to admit 
of turning an entire train. Three of the most used arrangements 
are given below. 

FiHST Casb. — One branch of Y a straiglii line. 

This is only the special case of the last problem in which the 
cutting line becomes tangent to both curves. In Fig. 56, if any 



A 




F 




c 


^ 






'■'V-- 


^^--^ ' 






'V, 


>v; \ 


'"n /^ 




E 


R 

_../ 


/ 


■^ 


U l,.^ es 


"y{ 






>? 


^1 


O 


l"-^ 




. 







Fia. 56. 

one of the points -4, B, or C is given, the others may be located 
by finding the angles c and 6. Draw OiE parallel to CA ; then 
in triangle OOxE 

(i? + ^i)cos6 = n- Bx, 



cos 6 = 



B-\-Bx 



(107) 



This follows at once from (103) by making angle a — 0. Then 
angle c = 180 — ft. If AB were a located curve and the point 
5 given, formula (107) would furnish us a value for i?.. 

AnotlieT solution is to prod\ice the tangent at B to cut AG Hi F; 
tlien AF= FC = BF, Join 2^ with and 0» ; it can easily be 
seen that angle OFOx = 90% and, by geometry. 



Therefore 



and 



BF= Y^X Bi. 
ton i« = -g- =1/5;. 



(108) 
(109) 

(liO) 



LOCATIOK. 



101 



Example.— Let AB be a 8* curve, £(7 a 6* curve, the point A 
at station 180. 
By (107), 

"^ ' = iSSrfSi = ^-^^^^ = ''' ''' ^^• 

The number of B is 180 + 23.511 = 203 + 51.1. Angle e = 
109** 28', equivalent to 18.244 stations on the 6** curve. 

Second Case.— I%« three brancTiee curbed and convex iotoardi 
eac?i other. 

Given the three radii and any 
one of the points A, B, or O, 
Pig. 57, we have only to find the 
angles at the center, then divide 
these angles by the degrees of 
the respective curves to get their 
lengths and locate the three 
branches. 

In the triangle 00x0%, letting 
00 1 = I, 0,0a = m, OOt = w, 
»= ta + wi-f-n) = /? + /?, +i?„ 
we shall have, by trigonometry. 




Pio.67. 



cos 4a - i/ ^" - Q _ Ab + STT^S. 



(111) 



Angles b and c may be found in like manner. 

The angles may be found otherwise by letting fall a perpen- 
dicular from one vertex upon the opposite side, as OE perpen- 
dicular to OiOs. Then from the relation 

OiOi : OiO + OOt = OOi - OOt : OxE - 0%E 



determine O^E and OiE\ then the right triangles O^OE and 
OiOE yield values of cosine a and cosine c, after which & may 
readily be obtained. 
Third Case. — One branch concave to the other two. 
In Fig. 58 the triangle OOj Oa may be solved for the angles at 
O, Oi , and 0« ; for if the radii are given, the sides OOi = R — Ri, 
00% = i? — -Ra, and 0,0i = i?i -|- iJ« are known and the solution 



102 A FIELD-MAKUAL FOR RAILROAD EKGIKEERS. 

is the same as for second case. Then b is the central angle for 
curve AB, a' = 180 — a, the central angle for AG, and c' = 
180 - c, the central angle for curve BO. 




Example.— If A is at sta. 830 on the 1** curve AB, AC an 
8" curve, connect with a 6" curve CB. Here we have 

OaO =^730- 717 = 5013, OiO = 5730 - 955 = 4775, 

and 0,0, = 955 + 717 = 1672. 

Solving this triangle, we get e = 88^ 20', b - 19* 28', and 
a = 72** 12'. The number of B is therefore 820 + 19.467 = 

889 + 46.7 ; the length of CB is .?1^ = 15.278 stations, and 

107 8 
of ^Ois -^^ = 13.475 stations, 
o 

131. To Locate a Reversed Curve between Parallel 
Tangents. 
First Case. — Badii equal. 

(a) The equal radii R and distance p between tangents known. 
In Fig. 59 draw 0^ parallel to -4 G^ to meet Oi 5 produced. 
From triangle OEOx, 

cos«=M^ = l-^. ("^) 

and 0^ =r 2JS sin a (118) 



r 



LOCATION. 



103 



From triangle ABO, 



AB = ~^— = p cosec \a = ^OE* + p*. . (lU) 




Fxo. 08. 

(5) ^6^ and p known, R required. 

Here AB = \/AQ^ -j- p« = *. Draw OJJ to the mid-point of 
AC. Triangles iL OiT and ^^(? are similar and AH ^ \k. 
Therefore 



1* 






whence 



4p 



(116) 



Example. — Connect two parallel tracks, 30 ft. c. to c. by a 7* 
reversed curve. From Table I, i? = 819 feet, and, bj (112), 

cos a = 1 - -^ = 0.98167 = cos 10* 59'. 



By (113), OE = 1688 X .19052 = 812.1 feet. 



By (114), AB = i^(312.1)» + (30)« = 313.6 feet. 

If p = 80, 0^ = 812.1, or AB = 318.5 Lad been given, we 
should have had, by (116) 



«.2^.„.,«. 



104 A FIELD-MANUAL FOR RAILEOAD ENGINKEBS. 

Second Case.— Badii unequal. 

(a) Suppose the radii B = OA and Ri •= 0,5 (Fig. 59) to be 
known We must find central angle a and AB = k. From the 
triangle OOiE, 

Then AB will be given by (114). 

{b) Suppose AB = k, p and B known, to find Bi and angle a. 

Triangle ABG yields 

sinla = -|- (117) 

0, LB is similar to A OB. Hence 

Bi^_ ± 
LB" p' 

But AC = 2B sin ia, and Z-B = i(k - ^C7) = 1(7«. Inserting 
this value of LB and solving for Ri, 

* = ^ <"«> 

From similar triangles, 



Inserting me value of Ci = -^^-^ from (118) and solving for 



Bi , we get 



i?i = j^ - jB. (119) 



Example.— -4J? = 800\ p = 30', i? = 819 ft., to find angle 
a and Bi. 

By (117), sin ia = ^ = 0.10000 = sin 5' 44'. 
Therefore angle a = 11* 28'. 
By {IW. Bx = ^^^ - 819 = 681 ft., an S'' 26' curve. 



LOCATION. 



105 



132. To Connect Two Parallel Tracks by a OroMOver com- 
posed oi two D" Ourres with a Oiven Length of Tangent 
betixreen Points of Contrary Flexure. 

In Fig/60 let AFQB be the re- 
quired crossover, FG=ly EB=p^ 
and OA = OB = R known; 
angle a and AE = x are re- 
quired. 

Draw OM parallel to AE to 
meet (/B produced ; draw also 
00 parallel and equal to FQ; 
join nud O'. From triangle 
OO'O, 

I 



tany=^^, 



. . (idO) 

2i? 




Fio. 80. 



00' = — - = 2R sec y = V^R^ + ^. 
cosy 9 w 



Then in triangle OO^M, 



Oa ""21? sec 1 






coay. 



(121) 



(122) 



Now knowing y and t, 

a = f - y (128) 

Next, X = OMr= Oa sin 2 = 21? sec y sin «. . (124) 

Example.— Given i) = 7" 80', p = 62 ft., I = 100 ft., to locate 
crossover when A is at sta. 86 + 20. 

By (120). 

log tan y = 2 - 3.18441 = 8.81659 = log tan 8" 44'. 
By (121), 

log Oa - 8.18441 - 9.99908 = 3.18633 = log 1632. 
By (122), 

log cos 2 = 8.16643 - 8.18533 = 9.98110 = log cos 16* 47'. 
By (123), 

a = 16* 47' - 8* 44' = 13" 8'. 
By (124), 

log X = 3.18583 + 9.46053 = 2.64586 = log 442.4. 



106 A FltLD-MAKtJAti FOR ftAItltOAD fiNGlKEfitlS. 

133. To Find the Radius of th« Ravened Ounre AFE, Fig. 
61, Oiven Angles / end 1\ and 

From the figure, 

i?taiil7'= CF. 
Adding, 
B(\AXk \1 + tan i/') = J?C7=x *. 




Fio. 61. 



Whence 



-B = : 



(125) 



tan i/ -I- tan i/' 

Example.— Given / = 10% i'= 20*, EC = 700 feet, to find R, 
700 



By (125), i? = 



0.08749 + 0.17688 



= 2658 ft., a 2' 9f curve. 



134. To Locate a Reversed Onrve between Fixed Points. 

In Fig. 62 let AB — k, and angles 1 and 1' be known. We 




Fig. 88. 

have to find R and the angles a and h. 

Draw (yo parallel to, and 00 and (yF perpendicular to, AB, 
Angle AOQ = / and BOF = /'. Then Oi^ = i? cos / and OF 
= RcosI'. Hence 

aG' = i?(cos/+co8J0. 
In triangle 000, 00' = 2R. Therefore 

i?(co8 /-r cos 2') cos /-f cos/' 



cos X = 

2R 2 

an expression from which R has disappeared. 



(126) 



LOCATION. 



107 



We now have a = 7 + a? and & = /'+«. 

To find i?we have AE+ EF -{- FB = k, 
or i? siu i + 3i? Bmx + B sin I' = k. 

Whence i? = . , . . t , ,, , .. . . (127) 

sin / -f- sill /' -f- 2 SID a; ^ ^ 

J.n^^A^ expression for i? can be found by drawing ^iVand BL 
perpendicular to 00', and BN parallel thereto. Then, since 
:ifBAN=:x, 

Bsina -\- Bsinb = k cos x. 

• • ^ - sina + sind <^^^ 

Example. — Take the example of the last problem, 

A; = 700, J =10% /' = 20\ 
By (126), 

cos x = J(0.98481 + 0.93969) = 0.96225 = cos 15* 48'. 

We now have a = 25** 48' and b = 35** 48'. 
700 X 0.96225 



By (128), B = ^ 



= 660.2 ft., an 8* 41' curve. 



0.43523 + 0.58496 

136. To Connect Two Divergent Tangents by a Reversed 
•Ourve. 

First Case. — Advancing towards the P.I. 

Given the mdii B and Bi , the angle / and AO = k, to find the 
liugles a and b (Fig. 63). 



> 

1 


"X/ 


0, 


— fL^A 


''^^^^-vt!* 


H 



JL i !Q 

Fia. 68. 
Draw 00 parallel to the tangent BG to meet OiB produced 
Then EF - BO = AF- AE. 

Therefore BO =^ IRco^l — k sin /, 



108 A FIELD-MANUAL FOR RAILROAD ENGINEERS. 

From triangle OOiO, 

. Rx+BO 7?, + 1? cos/- A; sin J ^,-^, 

^"^' = -s+ir-= irvs. • • <^^^) 

Then a = MOxN =5-7, O.Jf being parallel to OA. 

SecoND Case. — Receding from the P.J. 

In Fig. 63 we have BC — ki , angle /, 7?, and 7?, given, to find 
angles a and h. 

Produce OA to meet 0,X drawn parallel to GA. AL equals 
OiJf = OMco&L 

OxH = Ri " BB = Ri - ki tan 7. 

.\ AL = 0,M = {Ri - ki tan 7) coa /. 
Hence 

OZ = 7? -f (7?i - A;, tan 7) cos 7 = 7? + 72, cos 7 - *, sin /. 

From triangle OOiL, 



cos a = 



PL _ /? -f 7?. cos 7 ~ k. slh 7 

oo; ~" 72 + iS • 

ft = a + 7. 



(130) 



Evidently, 

136. To Change the P.R.C. so that Second Branch of 
Curve shall End in a Tangent Parallel to Terminal Tangent 
and Distant p therefrom. 

In Fig. 64 let MAB be the located curve, EN = p. We must 




determine the angle COA, after which the desired curve ACB 
may be located. 
Draw EOi and XO, parnllel to T^Fnnd NQ. 

EL = OxK = p. 



LOCATION. 



109 



Fiom triangles OOi'ffmd 00,Z, 

(R + Hi) cos 6 = (2J + i?i) cos a - p. 



cos b = cos a — 



Angle AOC = b ^ a. 



Ji+Bi" 



(181) 



137. To Find the Radius of a Curved Track. 

Measure any chord AB = 21, and mid-ordinate CE = M, 




Fio. (J5. 

Tbeu in the right triangle OAE (Fig. 66), 



-R = : 



2M 



(182) 



CHAPTER IV 



TRAN3ITI0N-CUBVE8, 



Article 11.— Theory op the Tranbition-cubvb. 



138. Elevation of Outer Rail on Curves. —To counteract the 
effect of ceutrifugal force on curves the outer rail must be 
elevated above the iuner one. It is shov^n in mechanics that the 
ceutrifugal force is 



F- 



32.16-8' 



where W is the weight, « the velocity in feet per second, 82. Itt 

an average value of the acceleration of gravity in feet per secoud 

per second, and R the.radlus iu feet. 
In Fig. 66 let tlie vertical HL represent W, the horizontal KH 

the ceutrifugal force, A.B the plane of the rails, and CB = 6 
the superelevation of outer rail. 
From similar triangles, 

Equate this value of F to that given 
above and solve for e, giving 

Fia.66. ^-32.162?- • • ^^^> 

The gauge AB should be greater on curves than on tangents 
to allow for flange clearance and the effect of a rigid wheel-base. 
AC = 4.9 feet is about the right value for the horizontal distance 
between centera of rail- heads for standard gauge. In formula 
(133) « is in feet per second, but the train velocity is usually given 
in miles per hour. Let V = velocity in miles i^er hour, then the 

110 




TRAK81T10K-CURVES. Ill 

22 

velocity in feet per second will be « = — - 7. lusertfng these 

15 

values in (133) gives 

4.9 X 484 F» F« , ..^.. 

" = 82.16X22522 =35' "^^"^5^- ' * ' <^^^ 

This elevation will be required from the P.C. to the P.T., but 
obviously it cannot be introduced suddenly, so that for easy 
riding the rate of increase of e should be uniform. From (134) it 
is seen that e varies inversely with if, which requires that when 
e = 0, M = infinity. Hence R must decrease from infinity to 
the radius of the circular curve, while e increases from to its 
maximum value. 

139. The True Transition-curve should satisfy formula (134), 
but so far no such curve has been found that will at the same 
time admit of the same ease of location as the simple circular 
curve. According to Rankine the first use of any other than 
the circular curve was made by Gravatt about 1828 or 1829, 
the curve employed being the curve of sines. Another method 
described by Rankine is attributed to William Froude about 
1842 ; this curve was worked up in the Engineering News by 
A. M. Wellington in 1890. Other approximations are the RaU- 
road Spiral, developed by W. H. Seailes in 1882, and the cubic 
parabola,. described by C. D. Jameson and E. W. Crellin in the 
Railroad and Engineering Journal^ 1889. 

In 1880 Elliot Holbrook described in the Railroad Gazette the 
true transition-curve applicable to small angles and short lengths 
of the curve. In 1893 C. L. Crandall published formulae and 
tables applicable to large central angles for both the offset and 
deflection methods. 

140. The Notation here employed will be explained with 
reference to Pig. 67. The curve CBB'C is the circular curve 
offset at (7and G' from the tangents by the amounts CiJand C'W. 
AGB and BG'A' are the transition-curves. A is theP.r.C, 
or point of transition -curve, G the P.O., B the P.G.j, ^' the 
P.TG.x, G' the P.T., and A' the P.T.i. The co-ordinates of G 
are AIT = x\ EG = y'\ of C, ar' and HG=F\ of B, AM = Xi 
and 2dB = y\. The length of curve from P. T. G. to any point P 
is I, and the whole length from P.T. G. to P C, is /». 



112 A FIELD-MANUAL FOR RAILROAD ENGINEERS. 



141. Equation of Transition-curve.— Since the rate of 
change of e must be uniform, (134) may be written 



e^H^ 



8p' 



(185) 




Fia. 67. 



in which k is the rate of rise of outer rail along curve, and p the 
varying radius of curvature. From the calculus pd<f> = d/, 
whence 



^ ~ dip' 



Insert this in (135) and solve for d<p. 



^k 



d<t> = y;}dl = 2mldl. 



(186) 



(137) 



2m is dependent upon V and k, and is constant for any one 
curve. 
Integrating (137), 

= mi«, . (188) 

the constant of integration being zero, for I is zero when is 
zero. 



TRANSITION-CURVEa 1 13 

From the elementary tiiangle drawn at the point P of transit 
tion-curve. Fig. 67, dl being tangent, 

f = «'" f- 

Expanding sin (p by trigonometry, 

hi which 3!=:3X2X1, 5! = 5X4X3X3X1. etc 
Substituting for <t> its value from (138), 

Integrating, 

Irat Tvt^* s 0, where ^ is in circular measure. To obtain <p in 
Q^gree3, <f> =^ ^'t^-^ = ^^. Inserting this in (139), 

^OU O/.o 

^^i(_V_ ^^ , 0°' 0"^ , ^ 

^ V171.89 79 X W ^ 8151 X 10^ 153245 X 10»« "T" • • -y » 
or y -IG» (140) 

in which G may be found from Table XIV with 0** as argument. 
Interpolation must be resorted to for values of not given in the 
table, or y computed by the formula. 
From the elementary triangle at P, Fig. 67 

dx 

■^ = cos 0. 

Expanding by trigonometry. 



114 A FIELD-MANUAL FOR RAILROAD ENG^iNEERS. 
Substituting mP for (p and integrating, 

^=V"'lO" + ^16"-936() + ---) ^^^^> 

Replacing ml* by <p reduced to degrees, 

_X 0^^' , 0'* 0°' , > 

^"~\^ "32828"'" 2328X10* 33114 X 10»» "^ * * 7 

or a; = i - ii?. (142) 

iZr varies with 0% and may be taken from Table XIV with 0* 
as argument. 

142. The Transition-curve Angle /j is the value 4> assumes 
attheP.C.i. From (188), 

7» = mli* (148) 

From (187) and (136), 

^_ _L = 
d<f> " 2ml ^' 

At the P. Ci p = if and may be taken equal to -^, so that 



whence 



J^ _ 5730 



1 7)** 



This value of «m In (148^ gives 

^'^h'^im (^^^> 

Reducing this to circular measure by writing ^»=^»**Igo=57^ 
gives 

143. The Oodrdinates of any point on the curve are given by 
(140) and (142). The length of the transitioD onrve beiu^j- knowu 



TRANSITION-CURVES. 115 

or assumed, ^i and Xi (the coordinates of the P.Ci) may be 
found from these equations by the help of Table XIV; the 
coordinates of the R 0, (see Fig. 67) will be 

F= y, — 72(1 — cos 7i) = yi — JB vers Ji, , . (147) 
«' = «, -i?8in/i (148) 

144. Deflection-angles.— With the transit at the P.T.G. (or 
P.T.i in backing up) the tangent of deflectionaugles may be 

found from the relation tan 5 = ?. Dividing (189) by (141), 

X 

tan 5 = -^ + .009523w«i« + .000167^'^^" + (149) 

o 

From trigonometry the expansion of the angle in terms of its 
tangent is 

6 = tan (5 - J tan^ d + J tan* 5 - etc. ... (a) 
In (149) write m^ = <f> and substitute in (a) : * 

<5 = ^ - .OO28230« - .000068^» (150) 

o 

From (188) and (143), 

/."^"?7""'* ^' 

in which -- = n. Prom (ft), <p = /inS4md this in (150) gives 
It 

8 = ^n* - .002823/iW - .0000687/n". . , . (c) 

o 

Both S and ii are in circular measure ; to reduce to degrt'cs 
multiply by t^. This gives, neglecting terms involving higlier 
powers of Ii than the third, 

6** = 4^ n«- .00000086 ii»n« (151)* 

o 

The second term is quite small, ftnd in most cases may be en- 
tirely neglected in puu :ice 



116 A FIELD-MANUAL FOR RAILROAD ENGIKEERS. 

With the instrument at any intermediate point a^V the deflec- 
tion-angle for any point xy, measured from initial tangent, will be 

tan S = ^ 2 ^. = i{w^* 4- ml"^ + rrdl") +t4?(w"^ + »»'^"*) 
+ ^{mHH" + m»«"»)-|- t«t(w»/*^"^ + mHn"*+m*i^n)+ . . . , (152) 

in which powers of mt^ higher than the third have been neglected. 
Substitute the value of tan S from (152) in (a), write mP = = 
/,7i', ml"^ = <p" = iif*"*, by (6), and reduce circular measure to 
degrees, giving 

S^ = 4-(»' + ^"* + ^^") - a small correction. (153) 
3 

For instrument at P.T.C, 7i" = ; then (153) yields 
(^0*) = "^^' ~ correction, 

o 

or 

(V) = ^\-5o (154) 

(154) is the same as (151), as it should be. 
For the transit at the quarter-point of transition-curve 

n" = r = T-' = 1 J then (153) yields 
*i Ci 4 

(*i°) = ^»' + A + i»») - correction. 



or 



(«i*) = ^'^ - Si (16^) 

For transit at midpoint of tiansition-curre n" = |, and, from 
(158). 

(8i*) = ^(»' + i + in) - correction, 



or 



(*0=3\-«i (168) 



TRANSITION-CURVBa 
For transit at three-quarter point vl* as | and 

( Y) = T<^* + A + W - corwclioii. 



117 



or 



For transit at P.(7.i n" = 1 and 



(1»7) 



(«,*») = :^V + 1 + n) - oorrecdoD, 



or 



(«.•) = ^-^A, - A. 



(168) 



With the transit at the P.T,G*i it will frequently be most con- 
venient to measure the deflections from the tangent to the circular 
curve at that point. Sometimes this will also be the case for the 
transit at the P.C.,. 

By reference to Fig. 68 it will be seen that for the transit at B 




the deflection from the tangent BC which serves to fix any point 
on the curve, as .6, is given by the equation 

or, in general, 



(V) = y^o + P,. 



(159) 



118 A FIELD-MANUAL FOR RAILROAD ENGINEERS. 

Table XV gives the values of A and B for the fi^^ positions of 
instrument for wbich equations (154) to (159), inclusive, were de. 

duced. The value of A must be multiplied by ~ , but B is taken 

o 

direct from the table in thousandths of a degree. 

If deflection-angles are wanted for other positions of the instru< 
ment, or for other points on the curve, they may be computed 
from equation (153). 

146. Tables. — Three tables are given for use with transition- 
curves. 

Table XIV was computed for use with formulas (140) and (142) 
in determining (7 and ^ ; being assumed and G and E com- 
puted. 

Table XV gives A and B for computing the deflection-angles 
by (154), (155), (156). (157), (158), and (159) for 20 equidistant 
stations on the transition-curve. For points not given in the 
table A and B must be interpolated. Linear interpolation will 
suffice in most cases, though when /i** is quite large second differ- 
ences may be preferable for A. B is given in the table in thou- 
sandths of a degree. 

Table XVI was calculated by assuming U in lengths varying 
by increments of 20 feet, then computing /,* by (146), yx by (139), 
Xx by (141), F by (147), and ic' by (148). y, and Xx will also be 
given more directly by (140) and (142) with the aid of Table XIV. 

The excess in length of transition-curve, measured from P, T.G. 
to the point on offset at P.C, over xf is tabulated as «; T is 
found by trial such that when inserted in (141) or (142) the same 
value of a;' will be obtained as in (148). This may be done by as- 
suming V a little less than -^ , then computing a;'. More than two 

trials will rarely be needed to find a sufficiently close value of V\ 
then « = Z' — a?', y' is found by (139) after finding l\ or <f)' 
may be found from (b) of 144, and used in (140) in connection 
with Table XIV. ^. - V is the length from O (Fig. 67) to the 
P.G.x\ the difference in length between this and the length of 
circular curve from P,G. to P. d is tabulated as «* ; that is, tf' = 
{Ix — V) — arc. Then e-\-e' =^ Ix — (j^ -\- circular arc). 

For values of h intermediate between those given in. the table 
linear interi^olntion will suffice, though second differences may 
be used for l^and y, if preferred. 



TRANSITION -CU RVES. 



119 



146. To Unite the Two Branches of a Compound Curve by 
a Transition-curve. 

The same objections hold to compound curves as to simple 
curves uniting with a tangent ; i.e., where there is a sudden 
change of curvature there should be a sudden cbange of super- 
elevation of outer rail, which of course is not allowable. Instead 
of compounding the curves, we may offset them at the P. C, C. 
and unite them by means of a portion of a transition-curve tangent 
to each of the simple curves. 

In Fig. 69 AB and CELM&tc the simple curves that are to be 
united by the transition-curve ANE. Extend the transition-curve 




Pig. 60. 

to (?, where its radius of curvature becomes infinite, and let 08 
be its tangent. Call the length of transition-curve from G to A 
li , from G to E Is, and from F to A h. E and A are points 
of tangency of simple and transition curves. Then Z, = i, — 1%. 
The coordinates of A arc Q8=Xi, 8A =y, ; and of V {WV 
perpendicular to G8), GW=Xi\ WV = F^ of E, OP = x^ , 
EP = 1/9; of L {LH perpendicular to 08), OH = Xz\ HL = Fz, 
Let BC=F^, 

The radius of curvature of transition-curve is inversely pro- 
portional to its length from G ; hence the curvature is propor- 
tional to the length of curve; therefore ^3 : ^1 = Dj : i>i , whence 



h = h 



A' 



(160) 



120 A FIELD-MANUAL FOR RAILROAD ENGINEERS. 
Then i. = it - ^. = h{l - §•) = ^i^^ST^'- • • (1«1) 

By (138) or (148). /! " (^ j * 

Equating the value of Js from this equation to that resulting 
from (146) gives 

/.-4^y=^-^* im 



WV= Fx and HL = Ft may be taken from Table XVI with 
/i and ;, as arguments. Then 0, W= i?i +1^» OtH^Rt+Fn 
Draw 0,r parallel to QS, then 0,r = IFiT; hence 

0»T=^(R, + F,) - (-Bi +1^»X 

and OiT »«/-»*'. 

Therefore 



OiOt = (a?/ - a?s') cosec a = Vo^ + Ti^. . 



(164) 



Then OT = CO, - 50,. 

or 2?; = iJ, - (i?i + OiO,). (165) 

The lengths of ^B and CE are 

^5= ^'';:^" 100, aw) 

027= '*'~^'' 100. (167) 

The excess of transition curve length over AB+ CEia 

e. = l,-{I^ + ^^)m. . . . (168) 



TRANSITION-CURVES. 



121 



It AB and CE are quite sharp, we must take account of the 
arc excess, so that we have then 

«, = ;,- M il^I± 4- ^^^^) 100 4- arc excess |. (168') 

'llie arc excess may be taken from the second column of 
Table IV, which gives the arc length for one station; this multi- 
plied by the number of stations gives the curve length, which 
may replace the values within the brackets in (168'). 

147. Ijength of Transition-curve to be Taken. — In practice 
the rate of change of superelevation of outer rail may vary from 

1200" *^ 400* ^*^^ ^^® ^*^ * ' ^^^^ evidently klj must equal the 
ijuperelevation of outer rail for circular curve ; or, by (185), 

F« 



Ml 



5730 



"Writing R = — ^i and solving for 1% 



Ix 



For k. 



1 

1200' 



17190* • 
Ix = 0.07 F*2). 



(169) 
(169') 



For k =5 ; 



For k = 



400' 



U = 0.085F»i>. (169") 

I, = 0.023 F«2). (169'") 



The following table gives values of ^i in feet per degree of 
circular curve for a few values of Fand k. 



k 


.30 Miles 
per Hour. 


85 Miles 
per Hour 


40 Miles 
per Hour. 


45 Miles 
per Hour. 


50 Miles 
per Hour. 


55 Miles 
per Hour. 


1 

-laoo 


63 


86 


113 


142 


176 


212 


1 

600 


8» 


48 


56 


71 


87 


106 


1 

m 


81 


29 


37 


47 


68 


70 



I 



122 A FIELD-MANUAL FOR HAILROAD ENGINEERS. 

When only a sliort tangent intervenes between two curves 
shorter ti-ansition curves must be taken, requiring larger values 
of k, so that overlapping may be prevented. 

For illustration suppose a 5** curve to be eased off with a tran- 
sition-curve, the highest train-speed being 45 miles per hour and 

* = sL- Sy tlie table the value of li will be 71 X 5 = 355 feet, 

so that we should probably take a 360-ft. transition-curve, re- 
quiring an offset of 4.7 feet by Table XYI. 

Article 12 —Field- work. 

A Field Formulas. 

148. For the cases most frequently presenting themselves in 
practice the foregoing formulas may be simplified so as to admit 
of the rapid location of points on the transition-curve with all the 
accuracy needed on location, though it is best to use the exact 
formulas and tables in setting track-centers on the finished road- 
bed. When the transition -curve angle is quite large it will be 
better to use the accurate methods on location also, but fur the 
more common cases the following formulas will answer. 

149. Simplified Formulas.— In (189) and (140) neglect, as 
small, all the terms following the first, giving 

y = '^' = |^=.OO5818Z0- ^^^^^' 

In (141) and (142) retain only the first two terms 

in which the last term is small for shoit transition -curves and 
may often be neglected, x being taken equal to /. 
The values of vn and I\ remain as before : 

_ 1 _ D ,.^
…[truncated]