….let's share our railways knowledge


By on May 12, 2013

Adhesion is one of the most important terms used in Railway parlance signifying the ability of efficient heavy haul and safe braking. Adhesion limits the tractive and braking effort per traction and braking wheel respectively. On Indian Railways,  adhesion limitation has caused more concern of wheel slipping and stalling of trains on gradient section and not much on issues of braking effort which is more common on European Railway system. Before defining Adhesion, it will be appropriate to understand the phenomenon of Rolling of the wheel.

Rolling of wheel


In fig 1 above, we all know Fr is very less as compared to Fs for the same weight but why and how much? This is because the coefficient of friction, in the case of rolling,  is very small as compared to sliding. Again why? For understanding this, it is important to calculate the speed at the point of contact when linear and rotational motion of the wheel is taking place. It is calculated by taking the vectorial sum of speeds at each point as given in Fig 2 above.

Let Linear and angular speed is V and ω respectively, with wheel radius r.  The vectorial sum of speed at the top and bottom is ωr+V and ωr-V. V is approximately equal to rω, therefore speed at the point of contact is almost zero unless there is a difference between V and rω. When relative motion between two surfaces is almost negligible, the theory of friction does not explain the phenomenon of what is happening at the point of contact. Rolling resistance is the term coined to account for the energy loss at the point of contact instead of rolling friction.

It is difficult to perceive almost nil speed at the point of contact. It can better be understood that if there is a rotational speed at the point of contact then it will be grinding the rail, a situation similar to wheel slip.

There are many theories given to explain the action taking place at the point of contact and summarized as follows:

Coulomb in 1781

  • Two famous Laws based on experimental studies
    • Resistance to rolling is directly proportional to the load
    • Resistance to rolling is inversely proportional to the diameter of the roller

Later studies showed these laws are only approximations but are used commonly for the first level of approximation

Reynolds (1876)

  • Rolling friction is due to the slip/creep taking place within the contact area

Heathcote (1921)

  • The surface speed of different points of a rolling sphere is different but travels the same distance on the surface.
  • Lubrication has a negligible effect on rolling resistance, but increases micro slip

Tomlinson (1929)

  • Molecular Theory of Friction: The rolling friction is due to the hysteresis in adhesion between the atoms of the roller and the surface. He suggested that when a roller rolls over a surface, the atoms are pulled from their equilibrium position, which on reaching a critical distance flick back to their original position. The hysteresis in this process leads to energy dissipation, resulting in rolling friction.


Bikerman (1949)

  • Surface roughness can have a significant effect on rolling friction only if it has a significant magnitude (height of hills) in comparison with the radius of the roller ( of the order of 0.1% of the roller radius)

Eldredge and Tabor (1955)

  • Main cause of rolling friction was the energy dissipation due to plastic deformations which reduces with each traversal. After a large number of traversals, the predominant reason for rolling friction becomes elastic hysteresis, since no plastic deformation takes places

Rolling resistance

Rolling resistance, thus accounts for hysteresis loss during elastic deformation and sliding friction during micro-slip of 0.1 to 0.2% [difference between rolling and linear speed (rω-v)]. When rw-V is high then the wheel is said to be slipping and sliding friction causes erosion of material from rail and energy loss. Rolling resistance is also expressed in terms of the factor (µr) which, when multiplied by its weight (W) gives the Force (Fr) requires to start the rolling action and is called the coefficient of rolling resistance on lines similar to the coefficient of friction.

Rolling resistance is very small in the order of 0.0001 as compared to sliding friction around 0.73-0.75 between two steel bodies.   This is how Physics explains why rolling is easier than sliding. Rolling resistance of railroad vs road is as follows:

Rail Road                                                      0.0001-.0002

Truck Tire on asphalt                              0.006-.01

Ordinary car tires on concrete            0.01-.015

Car tires on solid sand                             0.08

Car tire on loose sand                              0.2-0.4

This also explains the main reasons for the higher efficiency of rail transport as compared to the Road. Efficiency is slightly lost due to prevailing sliding friction between the flange and inner face of rail i.e. the factor ‘b’ which does not exists in road transport of the equation a+bv+cv2 and higher tare weight per unit carrying capacity which is not so in road transport.


It is important to note that in the case of traction wheel, the tractive effort is transferred to rail due to sliding friction in relation to micro-slip. Tractive effort vs. micro-slip characteristics is as given in Fig 3.



It may be seen that when the tractive effort is increased, slip increases in proportion. Maximum Tractive effort is attained when micro-slip is around 1.5% and is the maximum value of sliding friction which can be converted into adhesive weight responsible for the transfer of tractive effort. Thereafter, the capability of transferring TE reduces with an increase in micro-slip. Sliding friction in laboratory conditions between steel and steel is around 0.73 of which only a maximum of 0.42 is converted into adhesion due to practical surface conditions of rail and wheel. The coefficient of adhesion when multiplied by weight gives adhesive weight. Wheel slip starts when Tractive effort is more than adhesive weight.

Tractive effort control for best adhesion

It is not possible to attain maximum adhesion in Tap changer control system due to its inability to touch the peak adhesion and the likely occurrence of wheel slipping associated with auto-regression during progression in steps. This results in an average tractive effort being lower than the maximum adhesion available (Fig3). Tractive effort is reapplied manually in steps which again find difficult to touch the maximum level and therefore, maximum adhesion is limited to 0.35 in tap changer locomotive of WAG7 class of locomotive with high adhesion bogie. Zone A in the above figure demonstrates TE control in the Tap changer locomotive.

When tap changer locomotive works with auto-regression control isolated, the working will be in zone B where TE will get applied in excess to adhesion, with excess energy lost in heat due to sliding friction between rolling wheel and stationary rail.

With advanced power electronics and its control, it is possible to work at the maximum limit of adhesion (Zone C) where wheel slip is detected by comparing the sudden changes in acceleration and speed with reference values and corrective action taken to reduce tractive effort till such time acceleration and speed normalize. Condition of improved adhesion is sensed automatically and TE is restored without loss of time. A typical characteristic phenomenon happening is depicted in Fig 4 giving detection and correction cycle during the time period from O to C are in the following range


Fig 4

OA:      Detection of slip on any od the axle and slip condition recorded after 5msec.

AB:       Reduction of Traction motor current starts with a time constant of about 50msec.

BC:       1. The holding level of torque is maintained to stop the slip condition for a period of about 1sec.

              2.  Holding level is   decided by the inbuilt software program depending upon the severity of the slip

    3. When slip condition no more exists, Tractive Effort is reapplied with a time constant of 1sec.

This arrangement ensures the application of tractive effort to match the adhesion available (Ref. DMRC Project Re-adhesion)

Control of Traction Motor for best Adhesion

Three-phase Induction Motor:

For detection of wheel slip within 5msec means less than a quarter of one cycle rotation. This is not possible through the conventional method. Rotor current is produced by the same air gap flux, therefore, both are dependent on each other. In a three-phase induction motor, Torque dynamics will be limited by the dynamics of the flux, which has large time constant (300ms). To improve the torque dynamics, the torque and flux producing components of the stator current space vector isq and idq must be decoupled. On decoupling, the torque will have dynamics as that of the stator currents (10ms) which is generally faster by 20-50 times. De-coupling of the torque producing and the flux producing components of the stator current to achieve high dynamic performance in torque control is done by the method called Vector Control or Field Oriented Control. This control is similar to a separately excited DC series motor in which armature current and field are controlled separately.

The object of vector control or FOC is to separately control the magnitude of the two components of stator current called id(Flux controlling component) and iq (Torque controlling component). The approach is

Convert three-phase motor current into two-phase motor current by phase transformation. The current vector in the stationary frame is converted as iα and iβ. Based on an assumed position of the momentary rotor flux, this vector is decomposed into a flux generating component Id and a torque generating component Iq (Park d-q reference frame). The required voltage to the stator of the motor can be expressed in the same reference frame by means of the 2 dimensions of the vector: Usd and Usq.  These vectors are first transformed into the stator voltages Uα  and Uβ  and then again into three-phase control voltages. These voltages are the basis for a switching pattern to control the power electronics.

Separately Excited DC Series Motor:

WAG6A,B&C Thyristor control locomotives are provided with separately excited DC series motor, with slip-slide control where adhesion of 38% could be attained. The armature and field current can be controlled independently which are responsible for Torque and air gap field.

WAG9HI locomotive:

This is new development on the WAG9 class with higher weight and IGBT power devices is planned for provision of radar control to measure linear speed accurately as a reference speed to help in improving the adhesion further. There is individual control of motors and it is possible to control TE of each motor in case of drop in adhesion.

Weight Transfer:

As per BHEL documents (Ref 2), steady-state weight transfer between each bogie(L-95% & T-105%) and each axle (L-97% & T-103%) of the same bogie are worked out and tractive effort is controlled accordingly. This helps in applying tractive effort in relation to adhesive weight and reducing wheel slip probability. In tap changer locos, reduction of TE during start on leading wheels is reduced by field weakening but not very effective.

Dynamic Weight Transfer:

Weight transfer during start redistributes locomotive weight on each axle depending on the reactions and couples acting as mentioned in para 4.3. In case of a wheel slips in this condition, it momentarily disturbs weight transfer creating the condition of wheel slip for other axles also. With advanced microprocessor control, it is possible to identify the transient condition of weight transfer on each axle and reduce or increase the TE as required. (Ref.3 )

Review of Tractive control of Tap changer locomotive

Tap changer locomotive will continue to be with us for another 35 years, even if the manufacturing is likely to be stopped in a year or two. A review of the auto-regression control to contain the tractive effort within adhesive weight is required, which presently is based on cut in and cut out of relay Qd. During auto-regression, the holding current cannot be decided precisely but in steps. This can therefore be set at each one-step regression for a time period of 1 sec. This will prevent autoregression of more than one notch at a time, thus giving the early opportunity of using restored adhesive weight.


1. DMRC Project Re-adhesion

2. BHEL manual on Inverter Control

3. Development of Re-adhesion Control Method considering Axle-weight transfer of Electric Locomotive by Michihiro Yamashita QR of RTRI Vol 52 No. 1 Feb 2011

Post a Comment

Your email address will not be published. Required fields are marked *

There Are 3 Comments

  1. chandrasekhar says:


    I am very thankful to you for sharing your knowledge with all of us , I have recently gone through adhesion concept of yours which indicates that there no friction between wheel and rail

    My query was that if suppose we will not apply brakes to the wheels will there will be reduction in wheel dia if so why ?

    Thanking you sir

    • Mahesh Kumar Jain says:

      There is no friction between rail and rolling wheel. The loss of energy is due to rolling resistance and is described by many in their research work. One of it is due to molecular theory of hysteresis loss.I am unable to understand your question about reduction of wheel diameter. Wheel diameter cannot be reduced.

  2. 95Francisca says:

    Hi blogger, i must say you have high quality content here.

    Your page can go viral. You need initial traffic only. How to get it?
    Search for; Mertiso’s tips go viral