For official use only

GOVERNMENT OF INDIA
MINISTRY OF RAILWAYS

GUIDELINES FOR CARRYING OUT LOAD DEFLECTION

TEST OF STEEL PLATE & OPEN WEB AND
STEEL-CONCRETE COMPOSITE GIRDERS

REPORT NO. BS-128 (R1)

September, 2021

ISSUED BY

B&S DIRECTORATE
RESEARCH DESIGNS AND STANDARDS ORGANIZATION
LUCKNOW – 226011.

Page 1 of 13
Guidelines for carrying out Load Deflection Test of Steel Plate & Open Web
and Steel-Concrete Composite Girders

1.0 Preamble:

Load tests for PSC & RCC bridges are to be done as per provisions mentioned in Para
18 of IRS Concrete Bridge Code, 1997. For load test on Plate, Open Web & Composite
Girders, no provisions are available in the IRS codes. Generally load tests are not
required in Plate and Open Web Girders, but many times CRS, if considers necessary,
orders conducting of load deflection testing of steel bridges also as per provisions
mentioned in Para 17 of “The Railways (Opening for Public Carriage of Passengers)
Rules- 2000”.

The issue of load testing for steel girders was considered under Item 1065 of the 84 th
BSC held at Goa in November, 2016 wherein it was decided that RDSO should issue
detailed guidelines for carrying out load test for steel girders. Subsequently in 85th BSC
held at New Delhi in November 2018, the draft guidelines for load testing along with
proposed correction slip for amendment to the Railways (Opening for Public Carriage of
Passengers) Rules- 2000 were discussed.

This guideline is being issued for the purpose of carrying out load deflection test of
Steel Plate & Open Web and Steel-Concrete Composite Girders, if applicable and
considered necessary by CRS/CBE or any other competent authority. This guideline is
prepared for helping the field engineers. This guideline is just to facilitate and not to
supersede any of the provisions of the codes and manuals related to load testing of
steel bridge girder over Indian Railways. In case of any difference in this document and
any codes and manuals related to load testing of the steel girders over Indian Railways,
the provisions in the codes and manuals shall prevail over the provisions of this
guideline.

2.0 Applicability of load deflection test:

Load deflection test are normally and also historically not conducted in Steel Plate &
Open Web Girder Bridges and Steel-Concrete Composite Girder Bridges. Quality in
Page 2 of 13
fabrication & launching is ensured through strict Quality Assurance Procedures, at time
of fabrication of Jigs & Fixtures, approval of Welding Procedures, approval of Welders,
Inspection of Welding, Trial Assembly of 1st Span, Camber Values under Jacks &
without Jacks etc.

The applicability of the load deflection tests in case of steel Plate, Open Web and
Composite Girder Bridges will be as follows:

(a) For RDSO standard spans of Steel Plate & Open Web Girder Bridges and Steel-
Concrete Composite Girder Bridges, whose fabrication work is also inspected &
passed by RDSO officials or RDSO certified fabrication inspection unit under CBE of
Zonal Railways or any other third party inspecting agency being engaged with
approval of CBE of Zonal Railway, load deflection testing is not required. However, if
such type of girder is fabricated first time in Zonal Railway then the load deflection
testing may be done to validate the design and drawing.

(b) For non-standard spans of Steel Plate & Open Web Girder Bridges and Steel-
Concrete Composite Girder Bridges, if span is being used for the second and/or
subsequent time (span already under use in some other bridge over Zonal Railway)
and fabrication work is inspected and passed by RDSO officials or RDSO certified
fabrication inspection unit under CBE of Zonal Railways or any other third party
inspecting agency being engaged with approval of CBE of Zonal Railway, load
deflection testing is not required. However, if such type of girder is fabricated first
time in Zonal Railway then the load deflection testing shall be done to validate the
design and drawing.

(c) Load deflection test on Steel Plate & Open Web Girder Bridges and Steel-Concrete
Composite Girder Bridges as described in this guideline, may also be done as a
preliminary check on structures where there is doubt regarding serviceability or
strength of structure.

However, the applicability of the load deflection test as mentioned in (a) to (c) above is
for general guidance purpose. Load deflection test shall be conducted if applicable as

Page 3 of 13
 per the provisions of applicable codes and manuals in this regard or considered
necessary by CRS/CBE/Any other competent authority.

3.0 Test Train/Load for testing:

For load deflection testing, the test train should represent heaviest load which is likely to
run in the section. The test train should consist of One/Two locomotives along with
sufficient number of evenly loaded wagons with axle loads determined after weighing on
a weigh-bridge.

Effort should be there to cover the span fully with wagon load and apply maximum
number of axles of wagons on the bridge to generate maximum vertical load on the
bridge. Upto 30.5m span generally single loco with 5 loaded wagons is sufficient
whereas for 76.2m span double headed loco with 8 loaded Wagons will suffice the
purpose.

4.0 Test Speed:

If test train is rolled over the bridge at 5-10 kmph speed (Moving at dead slow speed),
results obtained can be taken for static condition. For dynamic tests, test train should be
moved at sectional speed. Speed should be noted precisely from speedometer in the
locomotive while passing over the bridge.

5.0 Measurement of Deflection:

Deflection can be measured with any of the following methods depending on the
suitability of bridge site, accuracy requirement and purpose etc. These methods are
explained briefly below:

(a) Card and Recording Pencil Method (Scale System):

Deflection reading is to be taken in the center line of track in mid-span of the Plate
Girder or mid-span of OWG. Arrangement should be such that test card is placed at
right angle to the center line of track (Arrangement should be independent of the
girders of bridge & preferably by use of CC Cribs with vertical arrangement coming
out of it) and recording pencil is placed/ tied to bridge, in alignment with center line
Page 4 of 13
 of track, brushing with the test card. Recording pencil point should be as fine as
possible.

(b) Dial Gauge Method:

Dial gauge should rest on independent footing of CC cribs and its plunger should
butt with soffit, at mid-span, in the center line of track. Selection of dial gauge should
be such that its capacity to record is more than expected deflection & wherein
pointer stops at the maximum deflection value. Card & Recording Pencil Method is
superior than Dial Gauge Method as it leaves permanent record of measurement,
whereas in Dial Gauge skill of BRI/person recording the reading becomes important
& also no permanent record is left.

(c) Linear Variable Differential Transformer (LVDT) method:

Deflection of the superstructure can be observed using LVDT (Linear variable
differential transformer), by erecting an independent mast on the non-yielding
support and with the magnetic base. LVDT having capability of recording
displacement up to 1/1000 mm connected with automatic data logger to record in
real time. This is one of the most accurate method for recording the deflection of
girder on real time with permanent records. However, it is difficult to use LVDT on
spans over water, road or with very high clearances.

(d) Laser image deflection measurement:

Laser image deflection measurement is realized by means of the good directionality
of laser. The laser fixed on the bridge creates a laser spot on the stationary
optoelectronic receiver, whose center moves equivalently as the bridge distorts at
different levels. Therefore, bridge deflection is available as long as the spot center is
obtained. When used on deflection measurement, laser image method is of high-
accuracy, which can achieve 0.1mm. Besides, its high sampling rate and low cost
are also suitable for bridges of small and medium size.

Page 5 of 13
 (e) Deflection measurement of bridge girder using Tort wire:
Mid span deflections of girder, can be measured by installing a tort wire between two
adjacent piers or pier and abutment. For this, an invar wire rope is used to prevent
changes from temperature. The wire rope is then kept tight with a counter balance,
as used for overhead electrical lines of Railways. A non-contact displacement
sensor is then used between the wire rope and the underside of the bridge deck.
The wire rope is considered the fixed point for reference which will work for normal
wind condition.

6.0 Permissible Deflection:

Permissible deflection should be calculated as per following steps:

(A) Note down design EUDL for Bending Moment from Bridge Rules say “X” &
corresponding impact factor say “𝑖”
Total LL with impact = EUDL(BM) + EUDL(BM) x (𝑖) = X (1+ 𝑖)

(B) Note down Dead Load (DL) which is self-weight/weight of fabrication of the bridge
(without bearings), from the bridge drawings. Calculate the total Superimposed
Dead Load (SIDL) consisting of weight of track structure (Rails, Sleepers & Fittings)
acting on complete span.

(C) Note down total design deflection value from bridge drawings (for Open Web Girders
total design deflection at mid span is equal to Camber Value provided). Steel is in
elastic state & deflection is directly proportional to the load applied. Total design load
consists of 4 parts i.e. DL+SIDL+Live Load “X” + Impact Load “X x 𝑖” as given in (A)
& (B) above.

(D) With total design deflection known and all 4 design loads known as given in (C)
above, calculate deflection due to Live Load say “LL” & deflection for Impact Load
say “𝑖” (values to be proportionally derived from total deflection value
corresponding to the load under consideration w.r.t. Total Design Load). If required,
deflection due to DL (“DL”) and SIDL (“SIDL”) can also be derived.
Total deflection due to LL = LL+i = LL + 𝑖
Page 6 of 13
 If design deflection values are not available in drawings, then same needs to be got
calculated through designer. Same procedure is to be followed in case of non-
standard drawings.

(E) Note down precisely axle loads & their spacing in the test train & with help of
Software “Clearance of Rolling Stock on Bridges” (The Software is available with all
Zonal Railways), calculate the EUDL(BM) for test train say (X’). Impact for Rolling
Speed (5 – 10 kmph) should be taken as “NIL” & for sectional speed impact should
be taken as 𝑖 ′

Sectional Speed
Where 𝑖 ′ = 𝑖 x .
125

Test train Live Load without Impact (at rolling speed) = X’

Test train Live Load with impact at sectional speed = X’ (1 + 𝑖 ′ )

(F) Calculate the Permissible Deflection as follows:

𝑋′
Permissible Deflection for Test Train without Impact (at rolling speed) = × 𝛿 LL
𝑋
𝑋 ′ (1+𝑖 ′ )
Permissible Deflection for Test Train at Sectional Speed = × LL+i
𝑋 (1+𝑖)

7.0 Example to calculate Permissible Deflection:

To understand para 6.0 an example is given below to find permissible deflection in case
of load testing of 30.5 m. span (31.9 m. Effective Span) MBG Standard Loading, welded
through type bridge to RDSO Drawing No. BA 11461-11571, with a test train of 2
WDG4 Loco plus 5 BOX‘N’ Wagons loaded to CC+6+2, running in the section at 75
kmph.

Permissible deflection should be calculated as per following steps:

7.1 Note down design EUDL for Bending Moment from Bridge Rules say “X” &
corresponding impact factor say “𝑖”

Page 7 of 13
 EUDL for BM for 30 m. = 278.1 tonnes & impact = 0.372

EUDL for BM for 32m = 293 tonnes & impact = 0.361

Derive X & 𝒊 for effective span of 31.9m

14.9
X = EUDL for BM for 31.9m = 278.1+ × 1.9 = 292.25 tonnes
2

0.011
𝒊= Impact factor for 31.9m span = 0.372 - × 1.9 = 0.36155
2

Impact Load = X x 𝒊 = 292.255 x 0.3615 = 105.65 tonnes

Total LL with impact = EUDL(BM) + EUDL(BM) X (𝑖) = X (1+ 𝒊) = 292.255 x 1.36155

= 397.9197 tonnes

7.2 Note down Dead Load (DL) which is self-weight/weight of fabrication of the bridge
(without bearings), from the bridge drawings. Calculate the total superimposed Dead
Load (SIDL) consisting of weight of track structure (Rails, Sleepers & Fittings) acting on
complete span.

From Drawing BA 11461 available on RDSO rail net Website.

Note Dead Load (Self Wt) from Drawing = DL = 55 tonnes.

For SIDL let there be 60 kg main line rail, 52 Kg guard rail & channel sleeper with
fittings @ 170 kg/sleeper at 1660 sleepers per KM density i.e. 170 x 1.66 =
282.2kg/m for sleeper & fittings.

SIDL = 60 kg/m + 82kg/m. + 282.2 kg/m. = 394.2kg/m say 400kg/m.

400
Super imposed Dead Load on bridge = SIDL = × 31.9𝑚
1000

= 12.76 tonnes

7.3 Note down total design deflection value from bridge drawings (for Open Web girder total
design deflection at mid span is equal to camber value provided). Steel is in elastic state
Page 8 of 13
 & deflection is directly proportional to the load applied. Total design load consists of 4
parts i.e. DL+SIDL+Live Load “X” + Impact Load “X x i” as given in para 8.1 and 8.2
above.

Note from Camber Drawing 11464/R

Maxm Camber Value at Mid Span = Design Deflection = 26.3mm.

This Deflection Value is caused by 4 Loads, calculated in para 8.1 and 8.2 above
which is mentioned below:

DL = 55 tonnes (i)

SIDL = 12.76 tonnes (ii)

LL (without impact) = X = 292.255 tonnes (iii)

Impact Load = X x 𝒊 = 292.255 x 0.3615 = 105.65 tonnes (iv)

Total Design Load = (i) + (ii) + (iiii) + (iv) = 465.679 tonnes.

Kindly see above 4 Loads causes total deflection/camber value of 26.3mm &
proportionate deflection due to each 4 components of load can be calculated on
proportionate basis.

7.4 With total design deflection known and all 4 design loads known as given in para 8.3
above, calculate deflection due to Live Load say “LL” & deflection for Impact Load say
“i” (values to be proportionally derived from total deflection value corresponding to the
load under consideration w.r.t. total Design Load). If required, deflection due to DL
(“DL”) and SIDL (“SIDL”) can also be derived.

Total deflection due to LL = LL+ i = LL + 𝑖

𝐿𝑖𝑣𝑒 𝐿𝑜𝑎𝑑
Deflection due to Live Load = LL = 𝑇𝑜𝑡𝑎𝑙 𝐿𝑜𝑎𝑑 × 𝑇𝑜𝑡𝑎𝑙 𝐷𝑒𝑓𝑙𝑒𝑐𝑡𝑖𝑜𝑛

292.255
= 465.679 × 26.3 = 16.51mm.

Page 9 of 13
 𝐼𝑚𝑝𝑎𝑐𝑡 𝐿𝑜𝑎𝑑
Deflection due to Impact Load =i= × 𝑇𝑜𝑡𝑎𝑙 𝐷𝑒𝑓𝑙𝑒𝑐𝑡𝑖𝑜𝑛
𝑇𝑜𝑡𝑎𝑙 𝐿𝑜𝑎𝑑

105.65
= 465.679 × 26.3 = 5.97 mm.

Deflection due to Live Load + Impact =LL+𝑖= 16.51 + 5.97

= 22.48 mm

Similarly DL & SIDL can be calculated.

7.5 Details of Test Train and calculations of Live Load EUDL (BM) from Software
“Clearance of Rolling Stock on Bridges” (The Software is available with all Zonal
Railways):-

Test train, as mentioned in para 8.0, has been assumed in present case to consist of
2 WDG4 Loco plus 5 BOX‘N’ wagons loaded to CC+6+2, running in the section at
75 kmph.

Note down precisely Axle Loads & their spacing in the test train & with help of
Software “Clearance of Rolling Stock on Bridges”, calculate the EUDL(BM) for test
train say (X’).

Impact for rolling speed (5 – 10 kmph) should be taken as “NIL” & for sectional
𝐒𝐞𝐜𝐭𝐢𝐨𝐧𝐚𝐥 𝐒𝐩𝐞𝐞𝐝
speed impact should be taken as 𝐢′ where 𝐢′ = 𝐱 .
𝟏𝟐𝟓

Test train Live Load without Impact (at rolling speed) = X’ = 277.6 tonnes.

Sectional Speed 75
Test train Live Load without Impact i′ = i X = 0.3615 x 125
125

= 0.2169

Test train Live Load with impact at sectional speed = X’ (1 + 𝑖 ′ )

= 277.6 x1.2169

= 337.81 tonnes.
Page 10 of 13
7.6 Calculate the permissible deflection as follows:

𝑋′
Permissible deflection for test train without Impact (at rolling speed) = × 𝛿 LL
𝑋
277.6
= 292.255 × 16.51

= 15.68mm

𝑋 ′ (1+𝑖 ′ )
Permissible deflection for test train at sectional speed = × LL+i
𝑋 (1+𝑖)

277.6 × 1.2169
= 292.255 × 1.3615 × 22.473

= 19.08mm.

8.0 Drawing of Conclusion from Load Deflection Test:

Normally deflection recorded should be less than the permissible deflection calculated
above. Reason for lesser deflection recorded is that steel structures have sufficient
residual strength beyond elastic limit and, in many structures, alternate load paths are
available and whenever load on structure exceeds the elastic limit of some part, the
same might not fully reflect in the deflection. Secondly due to simplifying assumptions
such as pin-jointed trusses, zero fixity at ends, ignoring the effect of track continuity and
2-D behavior of girders, the actual deflection of the structure is often lower than the
theoretical deflection computed.

In case, if deflection recorded is less than permissible deflection given in para 7.6
above, then bridge can be said to be in satisfactory condition.

In case, if deflection recorded exceeds the permissible deflection, but sum of recorded
𝐿
deflection plus theoretical deflection due to DL+SIDL is less than , then bridge is safe
600

from operations point of view but reasons for variations should be ascertained.

In case, if sum of recorded deflection plus theoretical deflection due to DL+SIDL

𝐿
exceeds600, then something is wrong and further detailed investigation of Bridge by

instrumentation or other suitable methods needs to be carried out.

Page 11 of 13
 It is important to understand that sometimes under Dead Loads, the deflection value
may be found abnormal. The possible reasons for abnormal values under Dead Load
are deformation/drooping of flanges during welding of plate girders, local deformation of
members during launching, creep & shrinkage effect of concrete in deck slab of
composite bridge etc. These factors may be kept in mind while evaluating the absolute
value of deformation. In load test, the differences of Initial deflection and the final
deflection i.e. the deflection under Live Loads shall be the important parameter to
decide the performance of the Girder.

9.0 Load Deflection Test of Steel-Concrete Composite Girders:

In Composite Girders due to composite bending action, tensile stresses are generated
in bottom steel flange & compressive forces are generated in concrete deck slab. The
shear connectors play the decisive role in composite action of the Composite Girder.
Being hidden item, inspection of shear connectors is not possible from outside. The
Load Deflection Test of Composite Girders should be done as per provisions mentioned
in Para 18 of IRS Concrete Bridge Code, 1997.

10.0 Limitations of load deflection test:

The various limitations associated with load deflection test are in case of steel girders
are given below:

(i) Steel is a perfect elastic & homogeneous material produced under strict quality
control over its chemical composition, physical & mechanical properties in factory
premises. The properties including ductility of steel as a building material are
normally beyond any doubt. The problems associated with fabrication, which are
often sought to be verified through load deflection tests, come into play only after
repeated application of loads (Fatigue Loading). One time application of load in load
deflection test is not the right tool for verifying the quality of fabrication.

(ii) Establishing independent reference to measure the deflection using dial gauge or
scale system is not possible at all locations and this means that the shore spans

Page 12 of 13
 where height is less and water not present are the often default choice for carrying
out the load test. This reduces the efficacy of load test.

(iii) Steel structures have sufficient residual strength beyond elastic limit and, in many
structures, alternate load paths are available. Even if load on structure exceeds the
elastic limit of some part, the same might not fully reflect in the deflection. Due to
simplifying assumptions such as pin-jointed trusses, zero fixity at ends, ignoring the
effect of track continuity and 2-D behavior of girders, the theoretical deflection
computed is often higher than actual deflection of the structure. Comparing the
actual field measurements of static load test with theoretical computations often lead
to erroneous conclusions.

(iv) For long span bridges, dead load and superimposed dead load itself is a substantial
component of the entire load carrying capacity. In this scenario, live load component
might not be significant for the overall girder (Though it is still important for some
individual components). For such girders, the camber values immediately after
launching and after providing superimposed dead load might give good idea of the
behavior of structure.

(v) The impact factor given in codes is a statistical value, which depends on several
factors such as condition of track, condition of vehicles, dynamic rail wheel
interaction and operation characteristics etc. which are difficult to create/replicate in
field. The error due to unknown impact factor might make the deflection reading
difficult to interpret.

***********************

Page 13 of 13