Welcome to Cecube Frequently Asked Questions. These pages address commonly asked traction related questions. They have a technical bias to varying degrees, dealing with electrical and mechanical matters related to traction drive vehicles such as regen brake and slip/slide control, as well as software and simulation. More general Railway FAQ regarding services, passenger issues and rolling stock class information are covered on numerous other sites. However, if your question is traction specific and is not addressed here please contact us.
TRACTION DRIVES - LIST of FAQ page 1
Q1. Why is there not a single simulation product suitable for all aspects of traction modelling?
The investment required to develop a robust numerical engine for a simulator is considerable. In the electrical / systems domain the MATLABTM and SPICE engines dominate the marketplace and most products involve developing front-end user interfaces (e.g SIMULINK) for these engines. The proprietary MATLAB engine is a powerful mathematical tool capable of solving general purpose high order differential equations. The application of non-linearities and equation constraints transform it into an application specific simulation tool. MATLAB is therefore very powerful when a problem can be described behaviourally or systematically, and it is multi-disciplinary. Via SIMULINK and dedicated optional tool-boxes the MATLAB engine will solve electrical circuit problems, but its strength is in multi-disciplinary system simulation and numerical analysis.
Conversely, the SPICE engine evolved with the primary purpose of modelling electronic circuits. It too has a powerful engine capable of resolving systems of differential equations. However, it is closely integrated to and optimised for commonly occurring constraints in electronic circuits. Application interfaces have expanded the capabilities of the SPICE engine to diverse fields by the use of behavioural modelling. This enables system modelling problems to be solved. However, its strength remains allied to electrical / electronic circuit descriptions, so would not be the simulator of choice for other domains.
From this background it is readily appreciated that the strengths of a particular simulation product depend on the originating goals of its numerical engine. Consider MATLAB ideal for multi-discipline system problems, whereas SPICE is first choice for predominantly electrical / electronic systems.
Q2. What do the acronyms MATLAB and SPICE stand for?
MATLAB stands for MATrix LABoratory and is now a product of Mathworks, Inc. It is at version 7. It was originally written in FORTRAN at University of New Mexico in the late seventies.
SPICE stands for "Simulation Program with Integrated Circuit Emphasis" written in FORTRAN at University of Berkeley, 1972. Latest and final version written in C around 1985 is SPICE3, which includes the analogue behavioural modelling capability.
Q3. What are the notches or gear changes heard on some AC drive units?
On some AC, PWM inverter drive systems (particularly GTO thyristor types) it is beneficial to keep the inverter switching frequency synchronous with applied induction motor frequency. As the motor speeds up so the switching frequency must increase as a multiple of the motor frequency. To prevent the switching frequency exceeding a practical upper limit, at predetermined speeds the switching frequency is dropped to a lower value, producing a new PWM mode. This mode changing is often referred to as "gear changing", from analogy with a manual gearbox and an ICE, making a distinct sound pattern that characterises the nature of the traction drive.
Q4. Some AC drive traction units do not appear to gear change. Why?
Most new inverter drive systems are based on IGBT (transistors) switching devices. Even high voltage IGBTs can now be switched at higher frequencies without excessive losses. By operating at a higher switching frequency (>1kHz) continuously, the PWM waveform can be asynchronous (not tied to the motor synchronous frequency) without undue additional distortion. This removes the need for a mode changing PWM strategy in order to keep the switching frequency synchronous.
Q5. What is "regen braking"?
The term "regen braking", or more formally regenerative braking, is the conversion of the train's kinetic energy to electrical energy by using the traction motors as generators. The electrical energy flows as current back to the supply system or into other loads. In principle this is easier with DC electrification than with AC electric railways, because the latter requires the phase and frequency of the generated electricity to be matched to that of the overhead line equipment (OLE). However, the fully-controllable front-end converters of a modern AC train perform regen braking reliably with accurate power factor control. The regenerated AC energy can return to the supplying grid and be used elsewhere. In practice, DC substations on older rail networks are not receptive, so energy saving is only achieved when an alternative load exists, such as an accelerating train, which can absorb the power. The regenerated AC voltage is in effect the train presenting a negative load to the OLE, which produces a slight rise in the system voltage. This results in a reduction in energy supplied by the generating stations on the grid network
The OLE is said to be receptive if a train or locomotive can use regen braking. If there are no other trains in the section that can absorb the power and if the substation is not designed to back feed into the supply grid, regen braking results in the OLE voltage rising above a predetermined limit, whereby the train control system detects the non-receptivity of the line. If the line is not receptive the train resorts to using friction or rheostatic electric brake.
Even if the line is receptive, feeding power back to the supply grid is not always possible because of practical constraints in the design of the substation equipment, such as fault detection in the 132kV supply system. Poor regenerated power factor control can result in false fault detection by supply distance protection relays. Regen braking is always most efficient in busy traffic sections. In a few railway systems the regenerated power is just dissipated using large resistive loads at the substation, negating the need to carry train borne rheostatic brake resistors to absorb energy. However, substation loads need to be switched in or out according to the direction of power flow, so are not often favoured by substation designers.
Apart from electricity cost savings for railway operators (usually 10% to 20%) there is an environmental and a carbon trading benefit. It also produces reductions in brake pad wear and the reliability of the whole mechanical braking system is improved. These are good reasons for railway systems worldwide to optimise regen braking strategies. This is analysed further in articles on electric regeneration.
Q6. What is unsprung mass and why is it important?
The unsprung mass is approximately defined as the mass between the rail / road and the suspension. This mass affects the natural frequency of the suspension and therefore the ride quality. Undulations in rail or road surface are impacted by a force proportional to the unsprung mass, whereas the sprung mass is isolated by the suspension. The impact causes rail / road forces that result in surface wear. This is particularly important at rail joints. On rail systems these track forces have to be minimised to prevent excessive track wear rates. These forces are vertical, lateral and longitudinal. The dynamic vertical force, or P2 force, is often used in assessment of design quality and acceptability.
Q7. What is weight transfer and why is it important?
When a powered vehicle, particularly a heavy locomotive, accelerates or brakes the distribution of weight across the axles shifts due to movement of the body weight centre of gravity. In conditions of low adhesion this can give rise to premature wheel slip or slide on one bogie. Modern slip control systems mitigate the effect but a good mechanical design reduces the weight transfer problem at source.
Q8. What is the difference between slip/slide control and creep control?
Both are electronic methods for preventing the complete loss of adhesion in adverse conditions. Slip/slide control refers to a system of comparing the wheel speed of a motored wheelset or bogie with the train speed. This is usually derived from a speed measurement on a trailer axle. A problem with this method is accuracy since wheel diameters of differing bogie will vary. By using many wheelsets, or by computationally estimating wheel diameters, the accuracy can be improved so that a deviation in speed of a powered axle is constrained by a reduction in the tractive effort delivered through it. This type of slip/slide control system is usually employed on EMUs and LRVs and gets vehicles through areas of poor adhesion.
For high power locomotives, in particular heavy laden freight, a system which optimises the available adhesion can produce better results. These control the extent of wheelspin to maintain the locomotive operating at the peak of the adhesion characteristic continually. This is achieved by analysing the delivered tractive effort in response to demand perturbations, to determine if the peak adhesion has been exceeded. If it has the mean demand in backed off slightly to maintain the operating point at the peak. To achieve this under time-varying shapes of adhesion curve, accurate speed measurement is important and may be provided by a compact Doppler radar targeted on approaching track sleepers. Such a system is displayed in the video example found here.
THIS SITE IS PROVIDED BY CECUBE ON AN INFORMATION ONLY BASIS. CECUBE PROVIDES NO WARRANTIES OR REPRESENTATIONS OF ANY KIND, EITHER EXPRESSED OR IMPLIED, AS TO THE CORRECT FUNCTIONAL OPERATION OF THE SITE OR THE INFORMATION CONTAINED THEREIN, THE CONTENT AND MATERIALS INCLUDED ON THIS SITE, INCLUDING WITHOUT LIMITATION TO THE ACCURACY OR COMPLETENESS OF INFORMATION. CECUBE LIMITED WILL NOT BE LIABLE FOR ANY DAMAGES OF ANY KIND ARISING FROM THE USE OF THIS SITE, INCLUDING BUT NOT LIMITED TO, DIRECT OR INDIRECT, CONSEQUENTIAL, PUNITIVE AND INCIDENTAL DAMAGES.