Long Trains test bench Following the American growing ... - UIC

Long Trains test bench
Following the American growing activity around the Electronically Controlled Pneumatic Brake (
ECP-Brake), in ’98 SAB-Wabco decided to start a research activity in the “long freight train”
control field, especially concerning the wired communication and brake management field.
Following this decision the R&D department in the Italian SAB-Wabco. S.p.a. Company begun the
study of the “wired” communication system and the design of a Long Train Test Bench: due to the
innovative research, a financing from the Italian Ministry of University and Research (MURST)
was requested and subsequently obtained.
At the same time, in order to participate to a bid announced by the railways DB AG and SNCF (for
the realisation of a 2250m. train, built up of 125 wagons and 16 locomotives, “wire bus” and “radio
bus” controlled) SAB-Wabco joined the French company ALSTOM Transport in a consortium.
This consortium, named SWAT, won the bid in ’99 and the new DB-SNCF project name became
FEBIS (Freight Electronic Brake Intelligent System). The FEBIS project takes today advantage of
the Italian SAB Wabco Test Bench for all the investigation and validation activities necessary
before the FEBIS components will be delivered to DB-SNCF for the final field installation.
For detailed references about FEBIS development, please see the article “FEBIS System aspects”
from S.Witte, R.Tione, C.Coulange, A.Launay presented in this session.
Overview
The purpose of the Long Train Test Bench LTTB is to allow the physical simulation of the
following subsystems:
The Pneumatic Brake subsystem
The Electronic subsystem
The “Double Wire” Communication and Energy subsystem
The Radio Communication subsystem
The above subsystems are referred to a train which max. parameters will be:
2250m. length
125 wagon max
4 locomotives
Due to the necessity to estimate the differences between a UIC train and a FEBIS train, and to
validate the FEBIS performances, the representativeness of the LTTB has been the major driver for
the LTTB design. Moreover, due to the huge quantity of tests to be carried out in different train
configuration, one of the major functional requirements for the LTTB design has been the high
flexibility in re-configuration.
The Pneumatic Brake subsystem
The pneumatic brake subsystem takes into account the necessity to reproduce freight trains,
characterised to have a Ø 1”_ brake pipe: the LTTB brake pipe has been built based on horizontal
loops of stainless steel pipes Ø 1”_ for a total length of 2284m.
Each loop is 45m. long: being the loops placed horizontally, the “train length development”
proceeds in the vertical direction: the mechanical structure sustaining this pipes envelop had to
grow on two floors, as can be seen on the following picture.

Long Train Test Bench “landscape”
Moreover, to allow simulation of “mixed brake pipe” trains, an additional brake pipe Ø 1” has been
added in the structure for a further length of 1246m, still developed with horizontal loops.
Both the brake pipes loops are built of linear segments 5m. long, linked together with different
connection types, which function is either to simulate the flexible hose inter-wagon connection,
either to allow the insertion of the wagons brake packages. This structure allows the creation of
different wagon lengths (modulo 5m.), as well as the creation of a variety of train compositions.
The brake package for each wagon is built of the following components:
brake cylinder with variable stroke
distributor valve with control reservoir
auxiliary reservoir
The following list of distributor and related components is available, for a total of 132 vehicles:
n° 45 SAB WABCO type W-U
n° 38 SAB WABCO type C3W
n° 29 Oerlikon type ESTGP
n° 4 Knorr type KE 2d SL-D
n° 4 Knorr type KE 2d SL-ALB/d8
n° 4 Knorr type KE 2d SL-ALD
n° 4 Knorr type KE 1d SL
n° 4 Knorr type KE Od KSL n 6d
The following list of Driver’s Brake Valve is available :
SAB-Wabco “Wabcotrol Unificato” with electronic control, developed and implemented on
“Trenitalia” specifications.

SAB-Wabco “Eurotrol” with electronic control, under SNCF homologated
SAB-Wabco PBL-91, SNCF standard
Oerlikon FV4, Trenitalia and SBB standard on old locomotives.
The Pneumatic Brake subsystem validation
Normally a train test bench is used to validate brake components in the standard UIC configuration.
In our case the LTTB is necessary to estimate and understand the real behaviour of a long train
during the braking phase, before this train will be physically built and moved on track, with all the
related risks.
For this reason the LTTB must be validated against real configurations in order to have a realistic
return from the simulations.
The DB and SNCF Railways have kindly supplied measurements done on real trains:
DB 650m. brake pipe Ø 1”_
SNCF 750m. brake pipe mixed configuration Ø 1” and Ø 1”_
Moreover other two references have been used:
SAB-Wabco France Train Test Bench 1200m. brake pipe Ø 1”_
ERRI B 126/RP 25 curve
A special mention must be given to the SAB-WABCO France Train Test Bench: differently from
the SAB-WABCO Italy LTTB, the French tool is built of real train component regarding flexible
interconnection hoses between wagons, compliant with the Fiche UIC 541-1, annex 4.
The SAB-WABCO France Train Test Bench composition respects the Fiche UIC 547, annexe 4:
this fact makes this device suitable to be a good reference for the LTTB validation.
The following picture shows a comparison between the DB train in Emergency brake and Service
brake.
LTTB – DB train comparison in Emergency

LTTB – DB train comparison in Service
The time difference between the two curves (anyway acceptable) is due to the different Driver’s
Brake Valves used on the different systems: The DB measurement is done using a DB locomotive,
while the LTTB measurements have been done using the SNCF approved PBL-91, which is
submitted to different requirements and is a little faster.
The next picture shows two couples of curves: one couple is a comparison between the SAB-Wabco
France TTB and the LTTB in the same configuration, the other coupe is the comparison between
the SNCF mixed configuration train and the LTTB equal configuration.
In this case the PBL-91 has been always used, and the superimposition of the curves is quite
evident, so confirming the full representativness of the LTTB.

LTTB – SNCF train comparison
The next picture shows the comparison between the two SAB-Wabco devices in 1200m. equal
configuration and the ERRI B 126/RP 25 curve.
LTTB – ERRI curve comparison
No reference is today available about a train 2250m. UIC brakes equipped, at least in a
configuration suitable to be compared with an equivalent SAB-Wabco Italy LTTB.

SNCF kindly supplied measurements about such a long train, but the composition was mixed Ø 1”
and Ø 1”_, in a unknown wagon configuration, so it was impossible to accurately reproduce the real
case.
The Data Acquisition System
The LTTB has been equipped with a Data Acquisition System DAS able to monitor and collect
pressure data during the simulations, developed and realised by SAB-Wabco.
The hardware characteristics of the DAS are:
200 pressure transducers installed on 50 wagons out of 132; the wagons are monitored in
four different points: the brake pipe, the auxiliary reservoir, the control reservoir , the brake
cylinder.
the pressure transducer precision is within ±1%the full range. An embedded calibration
process increases this precision
the sampling rate of the DAS is programmable up to 200Hz (5ms. acquisition period)
all the 200 transducers are acquired in less than 1ms. in order to have a negligible “time
skew” with respect the maximum sampling rate
The software characteristics of the DAS are:
recording and saving capabilities
replay, rewind and fast-forward functions
cursor are available during the replay function, for detailed measurements
Two kind of visualisations are available on the screen:
visualisation versus time
visualisation versus train length.
While the visualisation versus time is useful to verify the time dependence between different inputs,
the visualisation versus train length becomes relevant to observe and compare the different
pneumatic behaviour of the long train when in UIC mode and when in FEBIS mode.
The next picture shows an example of visualisation versus time, where two pressure transducers
waveforms are shown: in particular, the two pressure transducers are located on the Brake Pipe,
2153m. distant one each other, and the acquisition is performed during the propagation of the
accelerating chambers wave. The period between the two falling edges is 7.52s and evidences a
wave propagation speed of 286m/s.

Accelerating chambers wave propagation
The next picture shows an example of visualisation versus train length, where the four Pneumatic
Brake subsystem magnitudes are shown:
brake pipe (blue)
auxiliary reservoir (red)
control chamber (yellow)
cylinder (green)
The picture has been captured during a brake application in conventional UIC mode, 25s after the
brake application start (distributors in P mode).

UIC brake application after 25s
The Electronic and Communication Subsystem
The FEBIS Wired Communication system is based on the redundant Power Line concept. To be
able to properly simulate the reality, the double Power Line has been fully reproduced: horizontal
loops, made with the same type of cable used on the real train (shielded AWG8 cable with 50
characteristic impedance), have been placed in a receptacle running on the top of the LTTB
structure, in segments of 18m each, for a total length of 2500m per Power Line. Spare segments of
4m. each are available to create a huge variety of wagon lengths. Resistive series loads can be
inserted along the two Power Lines development in order to simulate exceptional voltage drops.
Obviously the two Power Lines can be added in series to measure the physical limits of the
communication channel.
Per each wagon two electronic Junction Boxes derivate the two Power Lines to the wagon
electronic equipment: a total of 132 Local Control Units LCU’s (this is the name of the wagon
electronic equipment in the FEBIS terminology) are available on the LTTB. The electronic devices
are hosted is racks placed around the LTTB, available for the developers investigations.

LCU’s installed on the LTTB
The two Power Lines are fed by a DC/DC static converter, delivering the 48Vdc and 230Vdc
requested by the system.
Finally, a Master Control Unit MCU (this is the name of the Locomotive electronic equipment in
the FEBIS terminology) has the complete control of the system.
A driver console allows the interface of the LTTB with the driver.
Concerning the Radio Communication channel, the LCU’s, as well as the MCU, will be equipped
with a 5.8GHz DSSS modem.
The LTTB is not provided of the axle generators, as these components are not directly involved in
the communication system.

Train dynamics simulation
One of the LTTB additional advantages is to produce reliable “real” data which can be used for
further numerical analysis.
In order to estimate the benefits brought by the FEBIS system in the “Long Trains” management,
the LTTB output data are used as input to train dynamics simulation tools.
While the author is writing, a research co-operation running between SAB-Wabco Italy, the Turin
“Politecnico” and the Rome University “Tor Vergata”, with the purpose to develop a Train
dynamics simulator which simulation level will include:
rail altimetrical profile
rail planimetric profile, in other words corners effect
wagons weight and different center of gravity position
wagons buffer to buffer reaction and geometry (not axial contact)
wagons suspensions reactions
rail to wheels reaction
friction material
In this moment he LTTB is equipped of a software “real time” train dynamics simulator, based on a
limited mathematical model.
The model describes a train where the mechanical links between wagons are assumed stiff: in other
words a heavy simplification has been done not taking in account the elastic “buffer to buffer”
various effects, neither taking in account the suspensions interactions.
This simulator anyway uses the altimetric profile applied on each wagon of the train consist;
moreover the shoes friction coefficient ( versus speed) is taken in account, as well as the railwheel
friction coefficient versus speed and versus planimetric profile.
Even if not at the level of the today’s most sophisticated simulators (e.g. E-Train), this one already
well helps giving preliminary train behaviours evaluations, waiting for the final one most advanced
developed with the Italian Universities.
Conclusions
The SAB-Wabco Long Train Test Bench represents today the European reference for the future
Long Freight Trains electronically controlled.
While the author is writing this article, the first FEBIS tests have been started on the complete
system: relevant results of this research will be available and shown during the WCRR presentation.
Just for completion a full FEBIS brake application on 2250m has been recorded and the result is
shown in the following picture.
The picture has been captured 2s after the brake command delivery from the MCU (FEBIS Brake
slope in P mode): it is interesting the comparison whit the picture “UIC brake application after
25s” previously shown.
It is evident the difference between the UIC mode and the FEBIS mode, in term of time delay of
application as well as in term of uniformity of brake effort applied along the train.
***FEBIS Brake application after 2s.***