F i i A f r r c L - t i f 0 3

M E A S U R E M E N T S O F R A I L / T I E

D E F L E C T I O N S A N D F A S T E N E R C L I P S T R A I N S

Facility for Accelerated Service Te/ting

TRANSPORTATION TEST CENTER PUEBLO, COLORADO 81001

This document is available to the public through The National Technical Information Service,

Springfield, Virginia 22161

PREPARED FORTHE FAST PROGRAM

AN INTERNATIONAL GOVERNMENT - INDUSTRY RESEARCH PROGRAM

U.S. DEPARTMENT OF TRANSPORTATION ASSOCIATION OF AMERICAN RAILROADSFEDERAL RAILROAD ADMINISTRATION 1920 L Street, N WWashington. D C 20590 Washington, D C. 20036RAILWAY PROGRESS INSTITUTE801 North Fairfax Street Alexandria. Virginia 22314

1-Track &lStructures

NOTICE

T h is document re p o rts events r e la t in g to te s t in g at the F a c i l i t y fo r A cce le ra te d S e rv ic e T e st in g (FAST) a t the Tran sp o rtatio n T e st Ce n te r, which may have re su lte d from c o n d it io n s , procedures, or the te s t environment p e c u lia r to th at f a c i l i t y . T h is document is d isse m i­nated fo r the FAST Program under the sponsorship of the U. S . Department of T ra n sp o rta t io n , the A s s o c ia t io n of American R a ilro a d s , and the Railw ay Progress In s t it u t e in the in te r e s t of in fo rm atio n exchange. The sponsors assume no l i a b i l i t y fo r i t s contents or use th e re o f.

NOTICE

The FAST Program does not endorse products or m anufacturers. Trade or m anufacturers' names appear herein s o le ly because they are considered e s s e n t ia l to the o b je ct of t h is re p o rt.

Technical Report Documentation Page

1. R epor t N o.

FRA/TTC-81/032. G overnment A c c e s s io n No. 3. R e c i p ie n t ' s C a ta lo g No.

4. T i t l e and S u b t i t le

Measurements o f R a il/ T ie D e fle c t io n s and Fastener C l ip S t r a in s a t the F a c i l i t y fo r A cce lerated S e rv ic e T e stin g

5. R epor t Date

O ctober, 19816. P e r fo rm in g O rg a n iz a t io n Code

8. P e r fo rm in g O rg a n iz a t io n R epor t No.7. A u t h o r s )

F ra n c is E. Dean9. P e r fo rm in g O rg a n iz a t io n Name and A dd re ss

B a tte l 1 e-Columbus La b o ra to rie s* 505 King Avenue Columbus, Ohio 43201

10. Work U n i t No. (T R A IS )

11. C o n t ra c t or Gra n t No.

DOT-FR-916213. T y p e o f R ep o r t and P e r io d Covered

In te rim Report O ctober, 1980

12. S p onsor ing A g e n c y Name and A d d re s s

U .S. Department o f T ra n sp o rta tio n Federal R a ilro a d A d m in istra tio n O ff ic e o f Research and Development W ashington, D.C. 20590

14. S ponsor ing A g e n c y Code

. RRD-3215. S up p lem e n ta ry N o te s

* Issu ed by the FAST program, T ra n sp o rta tio n T e st Center, Pueblo., Co lorado, Boeing S e rv ic e s In t e r n a t io n a l, In c . , O perations and Maintenance C o n tra cto r.

16. A b s t r a c t

Measurements o f r a i l - t o - t i e d e fle c t io n s and fa ste n e r c l ip s t r a in were made on both concrete and wood t ie t ra c k a t the F a c i l i t y fo r A cce le ra te d S e rv ic e T e st in g (FAST) in October 1980. These measurements were made to provide a data base fo r d u p lic a t in g , i f p o s s ib le , these d e fle c t io n s and s t r a in s in a la b o ra to ry . The la b o ra to ry te s ts would serve as the b a s is fo r e v a lu a tin g the performance o f va rio u s wood and concrete t ie fa ste n e rs .

The r e s u lt s demonstrated th a t r a i l - t o - t i e d e f le c t io n s are a reasonable way to c h a ra c te r iz e the fa ste n in g load environm ent. However, because o f the com plexity o f fa s te n e r d e f le c t io n , b e tte r r e s u lt s would have been obtained by d ir e c t measurement. The r a i l - t o - t i e d e f le c t io n s measurement d id p rovide enough data to develop ap p ro p riate la b o ra to ry fa ste n e r performance t e s t s .

17. K e y Words

Measurement Concrete t ie sR a i l / t ie d e f le c t io n s Wood t ie t ra c k Fasten er c l ip s S t r a in Measurement

18. D is t r i b u t io n Sta tement

T h is document a v a ila b le to the p u b lic through the N ational Te ch n ica l Inform ation S e rv ic e , S p r in g f ie ld , VA 22161

19. S e c u r i ty C la s s i f . ( o f th is report) 20. S e c u r i ty C la s s i f . (o f thi s page) 21. No. o f P ages 22. P r ic e

U n c la s s if ie d U n c la s s if ie d : 38: •!

Form DOT F 1700.7 ( 8 - 7 2 ) Reproduction of completed page authorized

1

PREFACE

T h is work was sponsored by the Improved Track S tru ctu re s Research D iv is io n o f the Federal R a ilro a d A d m in istra tio n (FRA ). Mr. Howard Moody of the FRA was the C o n tra ctin g O f f ic e r 's Techn ica l R e p re se n ta tiv e . Te st p lann ing and c o o r ­d in a tio n at the Tra n sp o rta tio n T e st Center (TTC) were completed by L a rry D a n ie ls , FAST Experiment M onitor. Members of the dynamic data c o lle c t io n group at TTC provided the data reco rd in g equipment and a s s is te d in the p la c e ­ment of the te s t f ix t u r e s on t ie s . The c o n tr ib u tio n of a l l of these people i s g re a t ly ap p re cia te d .

ACRONYMS

BCL B a tte l 1 e-Columbus La b o ra to rie s

DCDT d ire c t-c o u p le d d if fe r e n t ia l transform er

FAST F a c i l i t y fo r A cce le ra ted S e rv ic e T e st in g

FRA Federal R a ilro a d A d m in istratio n

IF in s id e r a i l , f ie ld p o sit io n

IG in s id e r a i l , gage p o s it io n

OF o u tsid e r a i l , f ie ld p o s it io n

OG o u tsid e r a i l , gage p o s it io n

T T C . T ran sp o rtatio n T e s t Center

ABBREVIATIONS AND METRIC EQUIVALENTS

0 degree = { ( °F -3 2 )5 / 9 }° C

11 ,* „ , in inch = 2.54 cm

kft kilohm

y in m icro inch = 2.54 x 10"8 m

% percent

lb pound = 0.45359 kg

sec second

V vol t

11T

TABLE OF CONTENTS

S e ctio n Page

EXECUTIVE SUMMARY. ......................................................................................................... .... . v i i

INTRODUCTION . . . . . . . . . .................. 1

TEST PROCEDURE .................................... ......................................................... . . . . . . 2

Approach ........................................................ 2Measurement L o c a t io n s ..................... 2In s t r u m e n t a t io n . ...................... 4C a lib r a t io n . . . . .......................... 4

TEST RESULTS . . . . . . . . . . . 7 . . . ........................... 7

C l ip S t ra in Due to In s t a l la t io n . .................................................. 7Dynamic Measurements .................................................................................................... 7Data Reduction ........................................................ 16

DEFLECTIONS AND STRAINS FOR LABORATORY TESTS . . . . . . . . . . . . . . 22

SUMMARY OF RESULTS . . . . . . . . . . ............................... . ............................... 24

CONCLUSIONS. ........................................................................... 26

APPENDIX A

L IS T OF FIGURES

F ig u re Page

1 P o s it io n s o f D isplacem ent Transducers .................. .... 3

2 Placement o f Transducers on the Four Types of Faste n e rs . . . . 5

3 E f fe c t o f Change in T ra in D ire c t io n on Maximum R a i l/ T ieD e fle c t io n s fo r Concrete T ie T ra ck ......................................... 8

4 Comparison o f T y p ic a l R e su lts fo r High and Low R a il onConcrete T ie T ra c k ................................................................................................ 9

5 Maximum R e s u lts from Hard and S o ft Pads on Concrete T ieT r a c k .......................... 10

v

L IS T OF FIGURES, CONTINUED

F ig u re Page

6 Maximum R e su lts from Measurements on Wood T ie s WithDouble-Shank E la s t ic C l ip S p i k e .......................... 11

7 Comparison of R e su lts fo r High and Low R a i ls on Wood T ie sWith Type A C l i p s ................................................................................................. 12

8 Maximum R a il/ T ie D e f le c t io n s fo r Wood T ie s With Type A C l ip s . . 13

9 E f fe c t o f Battered Weld on Spring-Loaded Displacem entTransducer............................................ 14

10 Transducer Arrangement and R a i l/ T ie D isplacem ents .................. .... . 17

11 C o rre la t io n of Measured and C a lc u la te d R a il Head L a te ra lD e fle c t io n . . ......................................................... 18

12 V e rt ic a l C l ip D e fle c t io n V s. C l ip S t ra in From TrackMeasurements and Two Lab o rato ry Te sts . . ...................... ..................... 19

13 Approximate Lo ca tio n s o f B ia x ia l S t r a in Gage P a ir s onInstrum ented Cl ip s . . ................................................ 21

A - l Schematic o f Type A C l ip In stru m en tatio n . . . ..........................................A-2

A-2 Measurement of C l ip Response V o ltage V s. D e fle c t io n inLaboratory F ix t u r e ........................................................................................ .... . A-4

L IS T OF TABLES

Table Page

1 Type A C l ip In s t a l la t io n S t r a in s .......................... 7

2 D e fle c t io n s and S t r a in s fo r Labo rato ry T e sts o f FastenersWith E la s t ic C l ip s ............................ 23

vi

EXECUTIVE SUMMARY

D uring October 1980, measurements of dynamic r a i l - t o - t i e d e fle c t io n s and fa ste n e r c l ip s t r a in s were made on fo u r se le cte d su bsectio n s of the F a c i l i t y fo r A cce le ra te d S e rv ic e T e stin g (FAST) t ra c k , Pueblo, Colorado. T e st lo c a ­t io n s were in 5° curves of wood and concrete t ie tra ck where e la s t ic c l ip fa ste n e r systems are in s t a l le d . R a i l - t o - t ie d e fle c t io n s were measured in the fo u r su b se ctio n s , and fa ste n e r c l ip s t r a in s were measured in a concrete and a wood t ie su bsectio n th at used a common c l ip . The measurements were to p rovide the data fo r la b o ra to ry s im u latio n of severe lo ad in g environments fo r the fa s te n e rs . Laboratory te s ts of fa ste n e r c l ip s t r a in v s . v e r t ic a l c l ip d e fle c t io n were made to compare with the f ie ld data.

The major r e s u lt s of t h is measurement program were:

• An approxim ate correspondence was found between the s t r a in -d e f le c t io n r e s u lt s from the tra ck data and the r e s u lt s produced by a la b o ra to ry method in which the v e r t ic a l lo ad in g was ap p lied through a r a i l segment.In another la b o ra to ry method, the v e r t ic a l lo ad in g was ap p lied d ir e c t ly to the c l i p . T h is method required much gre ater c l ip d e fle c t io n to produce s im i la r le v e ls o f c l ip s t r a in .

t The maximum le v e ls of c l ip s t ra in found on concrete t ie tra ck exceeded the fa t ig u e l im it id e n t if ie d by the m anufacturer in lo a d -d e fle c t io n te s ts s im i la r to those f i r s t t r ie d fo r t h is t e s t . T h is r e s u lt was e s ta b lish e d by a comparison o f c l ip s t r a in -d e f le c t io n data found in the f ie ld w ith the s t r a in -d e f le c t io n r e s u lt s o f the two la b o ra to ry te s ts and with the m an u factu rer's lo a d -d e fle c t io n data.

• Nonsymmetrical ro ck in g of the t ie about i t s lo n g itu d in a l a x is was found in most measurements on concrete t ie t ra c k .

• In c o n tra s t to the expected r e s u lt , maximum r a i l - t o - t i e d e fle c t io n s fo r co n crete t ie tra ck were found in a subsection co n ta in in g s t i f f r a i l pads ra th e r than in a neighboring subsection co n ta in in g r e s i l ie n t pads.

• For one type of e la s t ic c l ip wood t ie fa ste n e r, much of the v e r t ic a l r a i l / t i e d e fle c t io n occurred through t ie p la te bending rath e r than through f le x in g o f the c l ip .

The major co n c lu sio n s of t h is study are:

• The b a s ic approach was adequate to define lo ad in g environments fo r planned fa s te n e r fa t ig u e t e s t s . However^ the measurement s it e s were lim ite d in number, and data red uctio ns had to be performed by hand c a lc u la t io n s . In s p ite o f the l im it a t io n s , i t is reasonable to assume th at any response found d urin g t h is program w il l occur repeatedly during normal o p eratio n s on the FAST t r a c k . ,

vii

• Fasten er c l ip s in s t a l le d in tra ck re c e iv e s t r a in from sources independent o f v e r t ic a l d e f le c t io n . The la b o ra to ry te s ts in d ic a te th a t a s ig n if ic a n t a d d it io n a l s t r a in component i s produced by la t e r a l lo a d in g . Other p o s s ib le sources in clud e the lo n g itu d in a l r a i l / t i e motion and the s h i f t of the c l ip support p o in t a t i t s f la tte n e d toe due to t ie ro ck in g and r a i l r o llo v e r .

vii i

INTRODUCTION

D uring October 1980, measurements of r a i l - t o - t i e d e f le c t io n s and fa ste n e r c l ip s t r a in s were made on fo u r se le cte d subsections o f the F a c i l i t y fo r A cce le ra te d S e rv ic e T e s t in g (FAST) tra ck a t the T ra n sp o rta tio n T e s t Center (T T C ), in Pueblo, Colorado. A d d it io n a l la b o ra to ry te s ts o f fa s te n e r c l ip s t r a in v s . v e r t ic a l c l ip d e fle c t io n were made to compare w ith the f ie ld data.

The measurement program provided the data fo r la b o ra to ry s im u la tio n of severe lo a d in g environments encountered by re p re se n ta tiv e wood and concrete t ie fa ste n e r system s. T h is rep o rt summarizes the data and d e fin e s le v e ls of r a i l - t o - t i e d e f le c t io n s and c l ip s t r a in s to be used in fa ste n e r fa t ig u e t e s t s . The e f f o r t i s p a rt of a study to d efine improved fa ste n e r performance s p e c if ic a t io n s .^

FAST was se le cte d as a t e s t s i t e because i t p rovides an e a s i ly a c c e s s ib le tra ck and a severe fa ste n e r lo ad in g environment. Faste n e r performance problems have occurred in the 5° curves of both wood and concrete t ie t ra c k . These in c lu d e the f a l lo u t and f a i lu r e of fa ste n e r components on both types of t ra c k , t ie movement on concrete t ie t ra c k , and dynamic gage w idening on wood t ie t ra c k .

Measurement s it e s co n siste d of two su bsectio n s in the concrete t ie Sectio n 17 (5 ° cu rv e , 2% grade) and two su b se ctio n s in the wood t ie S e ctio n 07 (5 ° cu rve , e s s e n t ia l ly no g rad e ). A l l su bsectio n s contained "improved" fa ste n e r systems w ith e la s t ic c l ip s . The s ix r a i l - t o - t i e d e fle c t io n measurements con­s is t e d of three v e r t ic a l d e fle c t io n s a t the r a i l base, one la t e r a l a t the r a i l base, one la t e r a l a t the r a i l head, and one lo n g itu d in a l a t the r a i l base.C l ip s t r a in was a lso measured in one subsection of both wood and concrete t ie t r a c k , where a common c l ip was in s t a l le d .

The c l ip s t r a in v s . d e fle c t io n r e la t io n s h ip was measured in the la b o ra to ry by two methods: by v e r t ic a l lo ad in g of in d iv id u a l c l ip s in a sp e c ia l t e s t f i x ­tu re , and by lo ad in g through a r a i l segment to sim u late approxim ately the la t e r a l and v e r t ic a l load components experienced in t ra c k .

Track lo a d in g was provided by a sp e c ia l t e s t t r a in of two locom otives and 20 loaded 100-ton hopper c a rs .

' Dean, F .E ., Research Plan fo r the Development of Improved R a il Fastener Performance Reg uIrements, report by Batte I I e-Co Iumbus Laboratories to the Federal R a ilroad Adm in istration , Contract DOT-FR-1962, April 1980.

1

TEST PROCEDURE

APPROACH

The s e rv ic e lo ad in g environment of a fa ste n e r is d i f f i c u l t to measure in terms of the fo rce s to which a fa ste n e r is su b jected . Fastener load paths are complex, and space to in s t a l l tran sd u cers is u s u a lly very lim ite d .Measurement of w h e e l/ra il loads would not s u f f ic e because these loads are d is t r ib u te d to severa l ad jacen t t ie s , and i t i s d e s ira b le to r e s t r ic t la b o ra ­to ry f ix t u r e s to a s in g le fa ste n e r system . The s im p le st method by which the fa ste n e r lo ad in g environment can be defined i s to measure r a i l / t i e d e fle c t io n s and, where p r a c t ic a l , s t ra in s in fa ste n e r components. These can be reproduced in the la b o ra to ry to determine the fo rce s th a t sim u late the fa ste n e r lo ad in g environm ent. T h is method was adopted fo r the f ie ld measurements a t FAST.

The transducer lo c a t io n s shown in f ig u re 1 were se le cte d to provide measurements of a l l s ig n if ic a n t modes of r a i l - t o - t i e movement. The lo c a t io n s were:

• V e rt ic a l d e fle c t io n at the " f ie ld s id e le f t " p o s it io n ,• V e rt ic a l d e fle c t io n at the " f ie ld s id e r ig h t" p o s it io n ,• V e rt ic a l d e fle c t io n a t the "gage s id e le f t " p o s it io n ,« La te ra l d e fle c t io n a t the r a i l head,• La te ra l d e fle c t io n at the r a i l base, and• Lo n g itu d in a l d e fle c t io n at the f ie ld s id e base.

The three v e r t ic a l d e fle c t io n s can be combined to d efine c l ip d e f le c t io n s , r a i l / t i e r o llo v e r and r a i l / t i e ro ck in g . The d iffe re n c e of la t e r a l d e fle c t io n s a t the r a i l head and r a i l base can a lso d e fin e r a i l r o llo v e r . The only com­ponent th at was neglected was r a i l / t i e yaw. T h is could only be produced as the d iffe re n ce of very small l in e a r d e fle c t io n s in the la t e r a l or lo n g itu d in a l d ire c t io n s at the r a i l base.

MEASUREMENT LOCATIONS

Measurements were made in the 5° curves of concrete t ie S e ctio n 17 and wood t ie Se ctio n 07. The fo llo w in g su b se ctio n s were se le cte d fo r te s t in g :

T ie Type Fasten er Type T ie NumberConcrete Type B C l ip and S y n th e t ic Rubber Pad 0550, 056b,

(S o ft ) 0576Concrete Type A C l ip and P o lyeth y len e Pad 0390, 0405,

(Hard) 0415Wood Type A C l ip and Screw Sp ike s 0154, 0165,

0181Wood E la s t ic C l ip -S p ik e s and Cut Sp ikes 0339, 0351,

0363

To the extent th a t time allow ed, data were c o lle c te d on both r a i l seats of each t ie and fo r c lo ckw ise and co u n terclo ckw ise t r a in d ir e c t io n s . (Only the c lo ckw ise d ire c t io n was obtained fo r wood t ie t r a c k .) T ie lo c a t io n s were

2

FIGURE 1. POSITIONS OF DISPLACEMENT TRANSDUCERS.

3

se le cte d to be independent o f each o ther, w ith in the su b sectio n by a t le a s t 10 t ie s , and to o ffe r convenient f ix t u r e mounting. The concrete t ie t e s t segments were se le cte d because they o ffe re d t ie pads w ith w ide ly d if f e r in g s t if f n e s s e s . In the pad lo ad range between 4,000 lb and 20,000 lb , the s t i f f ­ness o f the po lyethylene pad is 4 .6 m il l io n pounds per in c h , w h ile th a t o f the sy n th e t ic rubber pad is 1 .5 m il l io n pounds per in ch . T ie pad s t if f n e s s c o n tro ls r a i l / t i e d e fle c t io n s more than any other param eter.

INSTRUMENTATION

A l l d e fle c t io n transd u cers were stan dard , d ire c t-c o u p le d d if f e r e n t ia l transform ers (D C D T 's). Except fo r the lo n g itu d in a l tra n sd u ce r, a l l were f i t t e d with preload sp rin g s to. e lim in a te the need fo r p o s it iv e attachment to the r a i l . The lo n g itu d in a l DCDT req u ired a ta rg e t th a t was clamped to the r a i l base. Ranges o f the transd u cers were:

• A l l v e r t ic a ls : ±0.10"

• Two la t e r a ls and the lo n g it u d in a l: ±0.5"

The v e r t ic a l transd u cers were se le cte d p r in c ip a l ly fo r t h e ir lim ite d p h y s ica l le n gth of about 3 .5 " , which m inim ized b a l la s t d istu rb a n ce . The l im it in f r e ­quency response of the DCDT's w ith preload s p rin g s was about 50 Hz.

To fu rth e r m inim ize the time required fo r setup, a transd u cer mounting f ix t u r e was co n stru cte d . F ix t u r e placement i s i l lu s t r a t e d in f ig u re 2. The f ix t u r e made i t p o ss ib le to t ra n s fe r the in strum en tatio n a rra y from one r a i l se a t to another w ith in 20 m inutes.

S tra in -g a g e d Type A c l ip s were a v a ila b le from an e a r l ie r study of the s t ru c tu ra l performance o f these c l i p s . 2 As shown in appendix A , a four-arm re s is ta n c e bridge co n siste d of a b ia x ia l p a ir o f s t r a in gages and two comple­t io n r e s is t o r s . The lo c a t io n fo r placement o f the gages on the c l ip s was se le cte d fo r convenience o f s t r a in d e f in it io n and is not the lo c a t io n where maximum s t r a in s can be expected. The c l ip s were used in these te s ts to pro­v id e a d ir e c t in d ic a t io n o f the lo ad in g environm ent of the c l i p , as opposed to the lo ad in g environment o f the fa ste n e r system th a t was provided by the d e fle c t io n measurements.

CALIBRATION

A ll transducer s ig n a ls rece ived supply v o lta g e (5V) and a m p lif ic a t io n from a m p lif ie r s . The s ig n a ls were fed to an 8-channel s t r ip ch a rt recorder in the TTC data van. F in a l gain, s e tt in g s were ad justed at the reco rd e r.

A l l d isplacem ents were c a lib ra te d by d e f le c t in g the transducer rod by a known amount and a d ju s t in g the recorder pen to a d e sire d sc a le p o s it io n .

Hadden,. J.H., et a l . , Tests fo r the Structura l Evaluation of Pandrol R a il Fasteners In Section 17 and FAST, report by Batte11 e-Co Iumbus Laboratories to the Federal Ra ilroad Adm in istration, Contract D0T-TSC-1595, March 1980.

4

(a ) Type A - Concrete T ie s (b) Type B Cl ip

(c ) Type A -

FIGURE 2.

Wood T ie s

PLACEMENT OF TRANSDUCERS(d) E la s t ic C lip -S p ik e

ON THE FOUR TYPES OF FASTENERS.

Transducers were c a lib ra te d fo r the fo llo w in g maximum p h y s ica l d e fle c t io n s :

• V e rt ic a l and r a i l head la t e r a l d e f le c t io n s : ±0.10"

• R a il base la t e r a l and lo n g itu d in a l d e f le c t io n s : ±0.05"

The instrum ented c l ip s were c a lib r a te d by shunt re s is ta n c e s placed across two opposite arms of the 4-arm b rid g e . Independent c a l ib r a t io n s were per­formed fo r measurement of s t a t ic s t r a in due to c l ip in s t a l la t io n and fo r dyna­mic s t r a in due to t ra in passage. T h is was d e s ira b le because the dynamic s t r a in was norm ally only a sm all f r a c t io n of the in s t a l la t io n s t r a in . The shunt r e s is ta n c e s , r e s u lta n t b rid g e o f fs e t v o lta g e s , and e q u iva le n t va lues of l in e a r c l ip s t ra in a t the gage lo c a t io n were:

Shunt R e sista n ce B ridge Response E q u iva le n tA cross Each o f 2 Arms V o ltage C lip S tra in .

In s t a l la t io nS t r a in

24.9 to ' 5V 3,330 y in / in

Dynamic S t ra in 100 kft 5 V 830 v in / in

A d e r iv a t io n o f the c l ip s t r a in -v o lt a g e r e la t io n s h ip s is g iven in appendix A.

6

TEST RESULTS

CLIP STRAIN DUE TO INSTALLATION

To i l lu s t r a t e the general le ve l o f dynamic c l ip s t ra in as compared to in s t a l la t io n s t r a in , measurements of in s t a l la t io n s t r a in were made at each d esign ated s i t e in the Type A C lip subsections p r io r to the s t a r t of dynamic measurements. Table 1 p resents the r e s u lt s of the in s t a l la t io n s t r a in measurements. Average le v e ls are la t e r compared with dynamic measurements.

TABLE 1. TYPE A CLIP INSTALLATION STRAINS.

C l ip Lo catio n C lip Type (V)Mean*

(u in / in )Range

(% - and +)

Concrete t ie s , low r a i l F ie ld c l ip 6.47 4,310 -5 , +4Gage c l ip 5.78 3,850 -9 , +3

Concrete t ie s , high r a i l F ie ld c l ip 5.43 3,620 -1 9 , +13Gage c l ip 6.00 4,000 -4 +6

Wood t ie s , low r a i l F ie ld c l ip 3.89 2,590 -9 +13- Gage c l ip 3.50 2,330 -5 , +5

Wood t ie s , high r a i l F ie ld c l ip 4.53 3,020 -5 , +9Gage c l ip 3.42 2,280 -1 3 , +23

*Three to f iv e measurements were made at each t ie lo c a t io n ( IG , IF , OG, OF) w ith in a su b se ctio n .

Note: The re la t io n s h ip between lin e a r s t ra in a t the gage lo c a t io n and tra n s­ducer vo ltage is e = 666e, where e = l in e a r s t r a in and e = transd u cer v o lta g e (see appendix A ). Mean l in e a r in s t a l la t io n s t r a in s fo r the two types of tra ck were:

Concrete t ie s : 3,940 y in / inWood t ie s : 2,560 u in / in

DYNAMIC MEASUREMENTS

Dynamic measurements were c o lle c te d during two n ig h ts o f o p eratio n s fo r a t le a s t one t r a in run a t each designated s i t e . The t e s t t ra in co n siste d of two locom otives and twenty 100-ton c a rs . There were three designated t ie lo ca ­t io n s per su b se ctio n . In the concrete t ie se c tio n , measurements were taken on in s id e and o utside r a i l s fo r both c lo ckw ise and cou nterclo ckw ise t ra in d ir e c t io n s . Lack of time r e s t r ic t e d the wood t ie measurements to the c lo c k ­w ise t r a in d ir e c t io n . Excerpts from some of the more in t e r e s t in g r e s u lt s can be seen in f ig u re s 3 through 9.

Maximum c l ip s t r a in s and r a i l / t i e d e fle c t io n s on concrete t ie s occurred on the low r a i l during co u nterclo ckw ise tra ve l (up the 2% grade) in the subsec­t io n w ith the hard pad. F ig u re 3 shows the e f fe c t o f t ra in d ir e c t io n ; f ig u re 4 shows t y p ic a l d iffe re n c e s between high and low r a i l ; and f ig u re 5 compares maximum r e s u lt s found on hard and s o ft pads. The la r g e r d e fle c t io n s occurred

7

„ 0- ir.CompKUi-con VERT, i FIELD 0 LEFT '

(LOW RAIL, HARD PAD, TIE 17-0550) .COUNTERCLOCKWISE TRAVEL

Lh±TT±5±3'TjZiZL±'i;i'T±T:L!:;.

CLOCKWISE TRAVEL

■ j1 sec.

. . . .

-■htd _L..L.i .1..;

Q ___ Compn.t&&2dn — T- • - rT ~ ’— p ! '"*■— j— t—

£ £ VERT.FIELD | o

« RIGHT h r

" ! " i ~.~i— r

■ r p , - { - t T : p _ ' i j r 1 . . p i

O.lpCompKeAiion

VERT, j o

LEFT h i - : ■ 1" ; ' ■ r ' ■

i" M - _Ul_L_

FIGURE 3. EFFECT OF CHANGE IN TRAIN DIRECTION ON MAXIMUM R A IL/T IE DEFLECTIONS FOR CONCRETE T IE TRACK.

8

FIGU

RE 4.

COMPARISON OF TYPICA

L RESU

LTS FOR HIGH AND

LOW RA

IL ON

CONCRETE TIE

TRACK.

LONG. RAIL/TIE LATERAL RAIL/TIE

£73TOm

cn

>X

2TOmoocrco■nTOOn:>TOa>acoo

"O>aco

ooooTOmm

to>o

LONG. RAIL/TIE

LATERAL RAIL/TIEDISPLACEMENT VERTICAL RAIL/TIE'DISPLACEMENT

FIGURE 6.

MAXIMUM RESULTS

FROM MEASUREMENTS

ON WOOD

TIES WITH

DOUBLE-SHANK ELASTIC

CLIP-SPIKE.

LONG. RAIL/TIE LATERAL RAIL/TIEDISPLACEMENT DISPLACEMENT

V E R T IC A L R A IL / T IE DISPLACEM EN T (in)r * o <

OoJ

oo50 -*l •< 3

r>or * n < 3m j> m T3 •-«»-* m *g m »-* m "q

H m —♦8O

o r n ? j . f » i r n p - H O * C*»

-r» m 7> ^£0*

$• P — 6 p —o S t—• O S»-. ° s -

FIGU

RE 7.

COMPARISON OF

RESULTS FOR HIGH AND LOW

R

AILS ON W

OOD TIES W

ITH TYPE A C

LIPS.

LONG. RAIL/TIE LATERAL RAIL/TIE

lo

ng

, r

ail

/t

ie

la

te

ra

l

ra

il/

tie

CL

IP

ST

RA

INS

(V

) '

DIS

PLA

CE

ME

NT

D

ISP

LAC

EM

EN

T

. V

ER

TIC

AL

RA

IL/T

IE,

FIGURE 8. MAXIMUM RAIL/TIE DEFLECTIONS FOR WOOD TIES WITH TYPE A CLIPS.

13

LON

G.

RA

IL/T

IE

LATE

RA

L R

AIL

/TIE

l D

ISP

LAC

EM

EN

T

: D

ISP

LAC

EM

EN

T V

ER

TIC

AL

RA

IL/T

IE

DIS

PLA

CE

ME

NT

0 .1C o m p A e - i i c o n

VERT, i FIELD 0 LEFT '

. -i-t — r-i-rr-... . 1 : i ; / . H / j . ’

1. 1 A C •

(LOW R AIL, T IE 1 7 -0 5 6 0 ) .

± : h : ± ; i; . , , . , 4 , ...... -

z?A5. se.y, •■ i rI i ; ! ■ — M L

r-H-r -i-t—r •1j_4-r-i.J _I. I :i : i • l I ■ M i . . :L £

Comp eiiioil VERT. I FIELD T 0 RIGHT

0 .1C om pA U iion

VERT, j ' GAGE 1 0 LEFT

“ l !j. •, ~ | I;'' •| i ‘ ■ v-.r i | i j 'M' . 7-t v i -r-f - H i m - n: I •!' ; \: - : i . ' I i -• i 1 3 i i ■■ , i :-ii- -i ■ ! ; : ' j 1 : 5 j !■■ *.1 • i ' : j -i. •: . 1 ; : ; _ 4 _ ' ! i _L -j_i— 1 * ■ ■ • ! ■ ! '■ r . ‘ 4 -» '...“ }— M -- i-- - -J— r-O—r- 11 4 -4 - U - : -L - l J j - T ! ■

-*— — 1— 1*VE " “ ■

-T.:; E l : i-r-1 ' r z z n o ! i ••! ' i t H E O u - 1 i - f i M - H - i - i• • _ . M ,1U - l '- L i- i . ! H i - i

"!*!?!: ::» jT'.hJ' • I' :!l ..!. ..1 : ! ; 1 1 ! i n I; 11 r : I •! 1 '! : ! ‘ ! ■ T T : 1 - ! ■ '! •'! - ■ i . f - ^ " p " :j •!'■ I-'L! .«•• i-H' L. H-ifti 1 ■•• •• i t ••. ; • • i ‘ 1 “ in------- j : . |- f 7" ;— L. .. 1 — r-\— rr-1— r - •— !— !— •. ; !— !— ;— . : ;

— J— -------- — . ; ! •;■-! l - . - M . r i ;■ ■ M i .--— - I M .— —* i . -T-.j. 1 . I'-'.{:! = ■; ■•!. : ; : •' i \

■ ' : : .v ii-^ ir: .f;* ;1. ::J :: " ; - i - 'i- i- 'i. -i.- if: i , ■ 1 .. : ' ! 1 i- :■ ■ i '! ’

: _i..

E E 1!

—i—rMil

U..L-

0 .1Outwasid I

RAIL ] HEAD 0 LATL.

• | • T- •! - , l > • -~ T i •• i I 1-;. I -T I i ' • ! r ~T

ywrX T l

FIELD LONG. 0

■ __c - j_ . _i-.t. j . — i. __ i_J_L-i-J.

rtT031!-r I - j:4-r— !-t ! t -

U pLL{i GAGECLIP 0

i

T " T l ". -I 1 1 r • ! • i ”1 '. ■ 1 ‘ « tT r r m r r - •’I-ti't ;-!-— ~I r t • /I-- !_i ! I ] : i- ‘V )_i !. ' I ■ j : ! ’ I :

H 4'i'*1"”' "■ "i'tr r- j j. T-L.j -j- , j. -: T~:- l j Z L Z J . ‘ !IH^LJlL:x.:|10:E : L'.i" i" J.LT-U1

U p l i t f 1

FIELD * CLIP

..l .i i

7:L:u:!!q:T T '■ {Tj” I •i - r i - n - ■ ' ' ; -r -:-T

!"r t t i t rr:r i-H-l-

j... t. | . i

FIGURE 9. EFFECT OF BATTERED WELD ON SPRING-LOADED DISPLACEMENT TRANSDUCER.

14

w ith the hard pads. However, the la t t e r comparison cannot be taken as an in d ic a t io n of the e f f e c t o f pad s t i f f n e s s , s in ce so few measurements were taken in both lo c a t io n s . V a r ia t io n s in load'due to p o s it io n on tra ck ap p aren tly outweighed the d iffe re n c e s due to pad s t i f f n e s s .

In most cases there was only one complete lo a d in g c y c le per tru c k , although sm all b lip s caused by the passage of in d iv id u a l"w h e e ls can occa­s io n a lly be seen in the v e r t ic a l d e fle c t io n measurements.

F ig u re 3 shows th a t , fo r th a t p a r t ic u la r su bsectio n and pad, in comparison to the c lo ckw ise t r a in run, the counterclockw ise run produced:

• Much h igh e r le v e ls o f r a i l head la t e r a l d e f le c t io n ,• Much h igher le v e ls o f gage c l ip s t r a in ,• Approxim ately equal le v e ls o f v e r t ic a l d e f le c t io n , and• Very small and approxim ately equal le v e ls of r a i l base la t e r a l d e f le c t io n .

T h e re fo re , most o f the r a i l head la t e r a l d e fle c t io n took p lace through r a i l r o llo v e r .

A t f i r s t g la n ce , the data in d ic a te ah anomaly, because la r g e r r o l lo v e r / la t e r a l d e f le c t io n and gage c l ip s t ra in should be accompanied by la r g e r v e r t ic a l gage d e f le c t io n . The anomaly is exp la in e d by exam ination of the phasing o f sim ultaneous f ie ld and gage v e r t ic a l d e f le c t io n s . These d e f le c t io n s are n e a rly "o u t-o f-p h ase" fo r the co u n terclo ckw ise t ra in run, and more n early " in -p h ase " fo r the clo ckw ise t ra in run. C a lc u la te d c l ip d e fle c ­t io n s presented in the next se ctio n show an approxim ate correspondence between c l ip s t r a in and d e f le c t io n .

Unsymmetrical t ie ro c k in g , or ro ta tio n about the lo n g itu d in a l a x is o f the t i e , was p re va le n t d urin g the concrete t ie measurements. T h is is made evident by la rg e d iffe re n c e s in the f ie ld s ide v e r t ic a l d e f le c t io n s measured at the l e f t and r ig h t faces of the t ie ( f ig u r e 3 ). As a consequence of t h is ro ta ­t io n , the c a lc u la te d c l ip d e fle c t io n s g e n e ra lly f a l l below the v e r t ic a l d e f le c t io n s measured a t the three p o s it io n s shown in f ig u re 1.

Measurements on wood t ie s w ith the double-shank e la s t ic c l ip - s p ik e s ( f ig u r e 6) show high v e r t ic a l r a i l / t i e d e fle c t io n s a t one transducer lo ca tio n and much lower le v e ls a t the other two lo c a t io n s . In both examples of f ig u re 6 , t h is is caused by t ie ro ck in g . In a d d it io n , the low r a i l example shows high r a i l head la t e r a l displacem ent with r a i l r o l lo v e r , and the high r a i l example shows la t e r a l d isplacem ent with almost no r o llo v e r .

L e v e ls o f v e r t ic a l d e fle c t io n measured on wood t ie s w ith Type A c l ip s exceeded the le v e ls measured on concrete t ie s w ith the same fa ste n e r by about a fa c to r o f two ( f ig u r e s 7 and 8 ) . In c o n tra st , the c l ip s t r a in s were much lower on wood t ie s . T h is is p o ss ib le because much of the v e r t ic a l r a i l / t i e d e fle c t io n takes p lace through t ie p la te bending. F le x in g of the t ie p la te , as i t conforms to the ir r e g u la r su rface of the wood t ie under com pression, can be observed as the wheel passes over the t ie . P la te bending was much lower w ith the e la s t ic s p ik e s , which are anchored to the t ie .

15

F ig u re 9 shows data taken on the low r a i l of the su b sectio n with hard pads, near a battered r a i l weld. The "spring-m ass" frequency response of the sp rin g -lo a d e d v e r t ic a l d isp lacem ent transd u cers is exceeded, producing apparent d e fle c t io n s up to 0 .0 8 0 ". Actual d e fle c t io n s were probably about0 .0 4 0 ".

DATA REDUCTION

The purpose of t h is f ie ld measurement program i s to d efine re p re se n ta tiv e , severe fa ste n e r lo ad in g environments th a t can be sim ulated in la b o ra to ry t e s t s . The t ra n s la t io n from f i e l d to la b o ra to ry measurements is made sim ple by the fa c t th at the same d e fle c t io n f ix t u r e and instrum ented c l ip s , c a lib r a te d id e n t ic a l ly , can be used in both ca se s. Upon v e r if ic a t io n of the d ata , i t is only necessary to s e le c t ap p ro p riate le v e ls of the measured v a r ia b le s to be reproduced under la b o ra to ry lo a d in g .

V e r if ic a t io n of the data was undertaken by two methods:

• The "rail" head la t e r a l- d e fle c t io n was "carcuTated from r a i l - base la t e r a l--------d e fle c t io n and the r o llo v e r obtained from f ie ld and gage v e r t ic a l d e fle c ­t io n s .

• The v e r t ic a l d e fle c t io n s a t the c l ip s were c a lc u la te d from the three measured v e r t ic a l d e f le c t io n s . Some correspondence between c l ip s t ra in and d e fle c t io n was expected.

F ig u re 10 i l lu s t r a t e s the p o s it iv e d ire c t io n s assumed fo r the measured v a r ia b le s and the ro ta tio n s re q u ir in g c a lc u la t io n . For a small number of maximum response cases se le cte d from the concrete t ie d ata , the fo llo w in g c a lc u la t io n s were performed:

R a il R o llo v e r A n gle___________ __________ _____

= Vfi ~ Vex 5.25R a il Head La te ra l D e fle c t io n

Gage C l ip D e fle ctio nV = l[ j + 6.12b x

F ie ld C l ip D e fle ctio nVgC = Vg"| + 0 .5 (V fr - V f i )

v fc - 0 .5 (vf| + V fr )

F ig u re 11 p lo ts measured v s . c a lc u la te d r a i l head la t e r a l d e f le c t io n . A syste m atic e rro r o f about 20% can be seen, as the c a lc u la te d d e fle c t io n s con­s is t e n t ly underestim ate the measured d e f le c t io n s . There are severa l p o s s ib i­l i t i e s fo r t h is e rro r: r id in g up of the transd u cer p o in te r on the r a i l head,ro ta tio n o f the d isplacem ent f ix t u r e due to t ie bending, or a s l ig h t m is a lin e - ment of the recorder pens. The la t t e r would produce nonsimultaneous va lues in the data se ts th at were read from the re co rd s, and m ight lead to underestim a­t io n of the c l ip d e f le c t io n s . Given these p o s s ib i l i t ie s fo r e r r o r , the degree o f c o rre la t io n shown in f ig u re 11 can be judged acceptab le fo r d e fin in g d e f le c t io n s and s t r a in s fo r la b o ra to ry t e s t s .

16

(b ) S im p l i f ie d R a il Segment

FIGURE 10. TRANSDUCER ARRANGEMENT AND R A IL/T IE DISPLACEMENTS.

17

DEF

LECT

ION

CA

LCU

LATE

D FR

OM1 R

OLL

AND

RBL

(i

n)

O.lOO

0.090

. 0.080

0.070

0.060

0.050

0.040

.0 .030

0.020

0.010

00 0.010 0.020 0.030 0.040 0.050 0.060 0.070 0.080 0.090 0.100

DIRECTLY MEASURED DEFLECTION ( in )

FIGURE 11. CORRELATION OF MEASURED AND CALCULATED RAIL HEAD LATERAL DEFLECTION.

18

6

5

4

3

2

1

0

1

2

3

.4

■5

•6

■7

V= 166y/in):

C L L p S & t a ln - V £ f t le .( L t io n {ftom

L c u t e A a l R u t ^ a l n t T e j> t

C ls L p S & ic u .n -V e .fiJte .c tA .o n T o j> t

.TRACK MEASUREMENTS

O Gage C l ip □ F ie ld Cl ip

■ 50 -40 -30 -20 -10 0 10 20 30 40

CLIP DEFLECTION (n)i;l )

50 60 71

IGURE 12. VERTICAL CLIP DEFLECTION VS CLIP STRAIN FROM TRACK MEASUREMENTS AND TWO LABORATORY TESTS.

19

F ig u re 12 p lo ts the c a lc u la te d c l ip d e f le c t io n s v s . c l ip s t r a in expressed in v o lt s o f s t r a in gage b ridge output. The eq u iva lences of s t r a in and vo ltage are stated on the f ig u re and d erived in appendix A. A lso shown on the f ig u re are the s t r a in -d e f le c t io n c h a r a c t e r is t ic s -o b ta in e d in the fo llo w in g two la b o ra to ry te s t s :

• V e r t ic a l c l ip s t r a in - d e f le c t io n . Each c l ip was in se rte d in d iv id u a l ly in to a sp e c ia l f ix t u r e and su bjected on ly to v e r t ic a l lo ad . B egin n ing w ith a s t a t ic load o f 1,700 lb , the c l ip was cy c le d between 1,400 and 2,000 lb . The re su lt .w a s a l in e a r and rep eatab le s t r a in -d e f le c t io n c h a r a c t e r is t ic th a t does not agree w ell w ith the f i e l d r e s u lt s .

• Sim ulated tra ck lo a d in g . A r a i l segment and fa ste n e r system were sub­je c te d to v e r t ic a l lo ad in g w ith the t e s t specimen mounted at

, la t e r a l/ v e r t ic a l angles from 25° to 3 0°. A reasonably complete s t r a in - d e fle c t io n c y c le was obtained fo r the gage c l i p , and the r e s u lt s were much c lo s e r to the tra ck data than the r e s u lt s from the v e r t ic a l load t e s t were. A complete c h a r a c t e r is t ic was not obtained fo r the f ie ld c l ip w ith in a v e r t ic a l load of 40,000 lb . However,- the p o rtio n of the ch arac­t e r i s t i c obtained in d ic a te s approxim ately the same s t r a in -d e f le c t io n behavior as th at fo r the gage c l ip .

The la c k of correspondence of the v e r t ic a l c l ip , s t r a in -d e f le c t io n t e s t w ith e ith e r the f ie ld data or the sim ulated tra c k lo ad in g in d ic a te s th a t the c l ip s in tra ck re ce iv e s t r a in from sources independent of v e r t ic a l d e f le c t io n . P o s s ib le sources of a d d it io n a l c l ip s t r a in in c lu d e lo n g itu d in a l r a i l / t i e d e f le c t io n , c lim b in g of the c l ip on the r a i l base from la t e r a l d e fle c t io n and r a i l r o llo v e r , and s h i f t o f the c l ip support p o in t at i t s f la tte n e d toe. The la t t e r could be caused by t ie ro ck in g and r a i l r o llo v e r .

The le v e ls o f s t ra in found in the f ie ld exceed the c l ip fa t ig u e l im it id e n t if ie d by the m anufacturer.2 In te s ts under v e r t ic a l load o n ly , the c l ip s d id not f a i l under dynamic d e f le c t io n s of 0.050" (over nominal s t a t ic lo a d ) , but f a i le d in le s s than 1 m il l io n c y c le s w ith dynamic d e fle c t io n s of 0 .0 6 0 ". T h e re fo re , a c l ip fa t ig u e l im it e x is t s a t s t r a in le v e ls produced by p u re ly v e r t ic a l d e fle c t io n s between 0.050" and 0 .0 6 0 ". To reproduce the s t r a in le v e ls found in the t ra c k , v e r t ic a l d e f le c t io n s up to 0.062" were req u ired in the la b o ra to ry f ix t u r e .

Many of the Type A c l ip s have f a i le d in the concrete t ie 5° curve o f the FAST t ra c k . Most of the f a i lu r e s have occurred at the in s id e -g a g e lo c a t io n on the t ie , and a t e ith e r of two p o s it io n s on the c l i p , as shown in f ig u re 13.The p o s it io n marked "S" i s the approxim ate lo c a t io n of maximum s t r a in , w h ile the p o s it io n marked "L" i s the lo c a t io n used fo r the f ie ld measurements. The peak dynamic s t r a in , the range of s t r a in (+ to - ) , and the mean s t r a in due to in s t a l la t io n a t p o s it io n "L" can be compared as fo llo w s:

Faste n e r C l ip S t r a in s , y in / inPeak Dynamic Dynamic S t r a in Mean In s t a l la t io n

S tra in Range, + to - S t ra in

Concrete T ie s 600 650 3,940Wood T ie s 250 280 2,560

^ H adden, e t a l . , o p . c l t .

20

" S !! Lo catio n •

Source: R ef. 2

FIGURE 13. APPROXIMATE LOCATIONS OF B IAXIAL STRAIN GAGE PAIRS ON INSTRUMENTED CLIP S.

21

DEFLECTION AND STRAINS FOR LABORATORY TEST

The f in a l products o f t h is measurement program are se ts of d e fle c t io n s and s t r a in s th a t w i l l su b je ct the fa ste n e r systems to r e a l i s t i c lo a d in g en viro n ­ments when ap p lie d in the la b o ra to ry . To d e fin e ap p ro p riate le v e ls , i t i s necessary to reco gn ize the l im it a t io n s under which these measurements were made. Perhaps the most im portant l im it a t io n was the shortness of the t e s t t r a in (2 lo co m o tives, 20 c a r s ) . The f u l l FAST c o n s is t produces high d ra ft lo ads in the 5° curve and 2% grade of the concrete t ie Se ctio n 17. There is evidence from e a r l ie r fa ste n e r measurements^ th a t a sh o rt c o n s is t does not produce e q u iva le n t d ra ft loads in t h is segment o f combined curve and grade.

In s p ite of the lim it a t io n s of the measurement program, i t i s reasonable to assume th a t any le v e l o f d e fle c t io n or s t r a in found in these measurements w i l l occur often under normal FAST o p e ra tio n s . Th erefo re , the le v e ls defined in ta b le -2 rep resen t envelopes over the maximum va lu es found in the measure­ments. Separate envelopes are defined fo r the wood and concrete t ie fa ste n e r system s, and they apply only to fa ste n e rs w ith e la s t ic c l ip s . The d e fle c t io n s and s t ra in s defined in ta b le 2 w i l l be reproduced in la b o ra to ry f ix t u r e s to d e fin e fa ste n e r fa t ig u e lo ad s.

Hadden, e t a l . , op. c l t .

22

TABLE 2. DEFLECTIONS AND STRAINS FOR LABORATORY TESTS OF FASTENERS WITH ELASTIC C L IP S .

R a il T ie D e fle c t io n sConcrete T ie

Fasteners ( in )

Wood T ie Fasteners

( in )

V e r t ic a l d e f le c t io n s a t c l ip s ( f ie ld s id e com pression or

gage s id e u p l i f t ) 0.040 *

Nonsymmetrical v e r t ic a l d e fle c t io n s

L e f t s id e of t ie * * ( f ie ld com pression or gage u p l i f t ) 0.020 0.050

R ig h t s id e o f t ie ( f ie ld com pression or gage u p l i f t ) 0.060 0.100

R a il head la t e r a l d e f le c t io n 0.10 outward 0.10 outward

R a il base la t e r a l d e f le c t io n ( a f t e r i n i t i a l lo a d c y c le ) 0 - 0.02 outward 0.10 outward

Lo n g itu d in a l d e f le c t io n 0.005 0.01

Type A C l ip S tra in * * * (V) 1 (V)

Peak s t r a in sGage c l ip ( u p l i f t ) 3 .6 (600 ye ) ' 1 .5 (250 ye)

F ie ld c l ip (com pression) 3 .6 (600 ye) 1 .5 (250 ye)

Range o f s t r a in s Gage c l ip (+ to - ) 3 .9 (650 ye) 1.7 (280 ye)

F ie ld c l ip (+ to - ) 3 .9 (650 ye) 1.7 (280 ye)

C lip d e f le c t io n s were not d e fin a b le from the t e s t s due to p la te bending. C l ip s t r a in data w i l l be used in ste a d .The d e s ig n a tio n s l e f t and r ig h t are in te rch an ge ab le .At p o s it io n " L " , f ig u re 13.

23

SUMMARY OF RESULTS

Fastener c l ip s t r a in -d e f le c t io n r e la t io n s h ip s were obtained from f ie ld measurements and from two la b o ra to ry t e s t methods. The second la b o ra to ry method showed an approximate correspondence with the f ie ld data, but the f i r s t method d id not. To produce the maximum le v e ls of c l ip s t r a in found in the f i e l d , the fo llo w in g v e r t ic a l r a i l / t i e d e fle c t io n s were re q u ire d .

V e r t ic a l D e fle c t io n At Maximum C l ip S tra in

( in )F ie ld Measurements 0.029 - 0.037

Lab Method 1 ( v e r t ic a l c l ip lo a d in g ) 0.062

Lab-Method 2 (sim u lated tra ck lo ad in g). 0.038

Dynamic c l ip s t ra in s were measured at one of the two lo c a t io n s on the c l ip where fra c tu re s have occurred in s e rv ic e . T h is was not the lo c a t io n of maxi­mum c l ip s t r a in . Maximum va lu es found at the measurement lo c a t io n were about 600 in / in fo r e ith e r u p l i f t o f the gage c l ip or compression of the f ie ld c l ip on the-low r a i l in the concrete t ie se c t io n .

The maximum dynamic c l ip s t r a in s found in the tra ck data exceeded the fa t ig u e l im it id e n t if ie d by the c l ip m anufacturer. T h is fa t ig u e l im it occurred at dynamic v e r t ic a l c l ip d e fle c t io n s between 0.050" and 0.060" in t e s t s s im ila r to the f i r s t method d iscu ssed p re v io u s ly . The reproduction of maximum dynamic s t ra in s found in the f ie ld req u ired 0.062" of d e fle c t io n in the la b o ra to ry t e s t by t h is method.

R a il h e a d -to -t ie la t e r a l d e f le c t io n s approached, but d id not exceed,~~0.10" on both wood and concrete t ie t ra c k .

Nonsymmetrical ro ck in g o f the t ie about i t s lo n g itu d in a l a x is could be observed in a m a jo rity o f the measurements on concrete t ie t ra c k . V e rt ic a l r a i l / t i e d e fle c t io n s measured a t opposite faces of the t ie were, in many c a se s , very la rg e a t one face (up to 0 .058") and very sm all a t the other fa ce , o ften w ith in ±0.006".

Sm all but c o n s is te n t lo n g itu d in a l r a i l - t o - t i e d e f le c t io n s up to 0.005" were measured.^

In co n tra st to the expected r e s u lt , maximum r a i l - t o - t i e d e fle c t io n s fo r concrete t ie tra ck were found in a subsection co n ta in in g hard pads ra th e r than in a subsection co n ta in in g s o f t pads. Maximum d e f le c t io n s on the hard pads exceeded those on the s o ft pads by about 30%. However, much of t h is d i f ­ference may have been caused by v a r ia t io n s in lo ad in g due to lo ca tio n on the t r a c k .

24

Measurements on wood t ie fa ste n e rs showed evidence of t ie p la te bending. Where the c l ip was anchored to the t ie p la te , most of the v e r t ic a l r a i l / t i e d e fle c t io n occurred through p la te bending rath e r than through f le x in g of the c l ip s . Although the c l ip of the second wood t ie fa ste n e r was anchored to the t ie , bending under compression could be observed in both systems as the t ie p la te s conformed to the ir r e g u la r t ie su rfa c e s .

25

CONCLUSIONS

T h is approach to measurement of r a i l - t o - t i e d e fle c t io n s and fa ste n e r c l ip s t r a in s i s adequate fo r d e fin in g the fa ste n e r lo ad in g environm ent. However, se ve ra l fa c to rs lim ite d the e f fe c t iv e n e s s o f the e f fo r t :

• Measurement accuracy would have been improved by d ir e c t measurement of c l ip d e fle c t io n s ra th e r than by the in d ir e c t method used.

• The t e s t s i t e s and t ra in d ir e c t io n s were se ve re ly lim ite d in number by reducing the n ig h ts o f t e s t in g from fo u r to two.

• The 2 0 -ca r t e s t t ra in probably produced lower loads on the 5° curve and 2% grade of Se ctio n 17 than t h e . f u l l FAST t r a in c o n s is t produces.

_._F.as.tener. lo ad in g environments _are c h a ra c te r ize d ^sep arate ly fo r wood and concrete t ie tra ck by the d e f in it io n of co n se rva tive envelopes over the m a x i-" mum d e f le c t io n s and s t r a in s found in t ra c k . In s p ite of the l im it a t io n s of the t e s t program, i t is reasonable to assume th a t any response found during t h is t e s t program w i l l occur rep ea te d ly in normal o p e ratio n s. Th erefo re , the d e f in it io n of co n se rva tive envelopes i s ap p ro p ria te .

Faste n e r c l ip s in s t a l le d in tra c k re ce iv e s t r a in from sources independent o f v e r t ic a l d e f le c t io n . The sim ple v e r t ic a l f le x u re of a c l ip in a la b o ra to ry f ix t u r e re q u ire s about 50% g re a te r d e f le c t io n to reproduce the le v e ls of s t r a in found at maximum d e f le c t io n s in the f ie ld or in a la b o ra to ry f ix t u r e th a t sim u lates f ie ld lo a d in g . The la b o ra to ry te s ts in d ic a te th a t a s i g n i f i ­can t component o f the c l ip s t r a in i s produced by la t e r a l lo a d in g . Other p o s s ib le sources of a d d it io n a l s t r a in in c lu d e lo n g itu d in a l r a i l / t i e motions and the s h i f t o f the c l ip support p o in t a t i t s f la tte n e d toe due to t ie ro ck in g and r a i l r o llo v e r .

26

APPENDIX A

STRAIN-VOLTAGE RELATIONSHIP OF INSTRUMENTED CLIPS

Figure A-l shows the schematic of the 4-arm bridge containing two active gages and two completion resistors. Changes in resistance of the active gages produce response voltage e0 according to the equation,

^o = I _ ^ 2 (A_1}e ” 4 R " R

where,e0 = output voltage e = input voltage R = resistance.

The linear strain at each gage is related to the fractional change in resistance of the gage by,

R1R

GF E 2

where,ei = linear strain at Gage 1 ^2 = linear strain at Gage 2.

(A-2)

A ratio between ei and eg was established in laboratory tests conducted in preparation for the field measurements described by Hadden, et al.l For biaxial gages placed at the position indicated in figure A-l, it was con­sistently found that,

£2 = -0.44 e1# (A-3)

Thus an effective Poisson ratio of 0.44 was found. Substitution of equations A-2 and A-3 into equation A-l yields,

— = — (1.44) ei- (A-4)e 4

Shunt resistance, Rc, placed across two opposite arms of the bridge will produce -the following ratio of response to excitation voltage:

£o = 1 R e 2 R + Rc*

R = 12CK2, nominal for a l l arms of bridge.

' Madden, J.H ., e+ a l . , Tests for the Structura l Evaluation of Pandrol R a il Fasteners In Section 17 and FAST, Report by Battel I e-Columbus Laboratories to the Federal R a ilroad Adm in istration , Contract DOT-TSC-1595, March 1980.

A-l

Red

FIGURE A-l. SCHEMATIC OF TYPE A CLIP INSTRUMENTATION.

A - 2

Amplification of e0 yields,

where,

Eo _ „ eo _ K R e " N e 2 R + Rc

(A-5)

E0 = amplified output K = gain factor.

Because the installation strain and dynamic strain were of such different magnitudes, two separate shunt resistances were used. To determine K in each case, the amplification was adjusted so that,

e 5

Then the two amplification factors were determined as follows:

a. Installation Strain: Rc = 24,900

Kinst.= 2(120 + 24,900) = 4 n

120

b. Dynamic Strain: Rc = 100*000

Kdyn ■ » - M ) . lj670.

Finally, the linear strain at Gage 1 can be expressed in terms of output voltage as,

e 1 = ----— ------- E1 K GF (1.44) e o

a. Installation Strain:

£ 1 (417)(2)(1.44)(5) E°= 0.000666 E0 (in/in)

= 666 E0 (uin/in).

b. Dynamic Strain:

£ 1 (1,670)(2)(1.44)(5) E°= 0.000166 Eq (in/in)

= 166 E0 (yin/in).

A - 3

CLIP DE

FLEC

TION

(f

n)

■ CL

IP DE

FLEC

TION

(j

n)/

CLIP OUTPUT (V )

-2 -1 0 1 2 -1 0 1 2CLIP OUTPUT (V)

FIGURE A-2. MEASUREMENT OF CLIP RESPONSE VOLTAGE VS. DEFLECTION IN LABORATORY FIXTURE.

A-4

To determine the relationship between strain gage bridge output and ver­tical clip deflection, the clips were subjected to the laboratory tests illustrated in figure A-2. An arbitrary strain "zero" was established with a vertical clip load of 1,700 lb. The load was then cycled between 1,400 lb and2,000 lb. The slope of the curve of strain (V) vs. deflection (in) was reaso­nably linear and very consistent, as shown in the figure. The mean slope of the two clips illustrated is 0.0173", when the clips are given shunt calibra­tions equivalent to those applied for dynamic measurements in the field.

A-5

Measurements of Rail/Tie Defelctions and Fastener Clip Strains, 1981 Francis E Dean

P R O P E R T Y O F F R A E E S E A E C E & B & V E L O P M E H

L I B R A R Y