Background
On railways an unpleasant high pitch noise can be generated from the wheel/rail interface as the train proceeds around curves. As the train moves through a curve the wheel pairs (which are fixed) travel different distances, and thus some sliding between the wheels and rails is inevitable. It is believed that the high-pitched squeal noise (in the region of 200 – 2000Hz) occurs at the leading inner wheels due to a lateral slipping at its natural mode [1]. When the friction characteristics of the wheel/rail interface allow it, the wheels can enter this unsteady dynamic. The unsteady dynamic is due to the wheels alternating between two sliding speeds, generating vibration. The vibration causses oscillation at the wheel web [2].
Creep Curves
We use creep curves to evaluate the performance of TOR materials and flange products. The test consists of ball and disc loaded against each other and driven independently, while measuring the friction force acting between them. The linear speeds at the point of contact of both the ball and disc can be varied from being equal to a difference of 10%. This difference in speed introduces sliding into the contact and produces the creep curves. The height of the creep curve gives an indication of the traction available for the train, while the shape of the creep curve can give an indication of noise.
A representation of a creep curve is shown in Figure 1. At lower levels of creep the traction increases linearly. This linear zone is due to more and more of the contact become sliding as opposed to rolling. At a certain percentage creep the contact becomes fully sliding, sometimes referred to as “saturated”. The traction forces are then dependent on the metal surfaces and any third body material. Three possible scenarios are shown in Figure 1:
- Curve “A” shows a traction force falling with increasing creep. Where the creep curve has this negative gradient, the wheels can enter an unsteady dynamic, where the speeds can quickly alternate between the two points of same traction either side of the saturation point. This produces a highly undesirable stick slip cycle, which will lead to the development of vibration and ultimately noise.
- Curve “B” will suppress the noise, due to the removal of the negative damping effect of the system depicted in Curve “A”.
- Curve “C” depicts the ideal system for a TOR material with high overall traction and a positive slope after the saturation point.
Results
Eleven railway products were tested using the creep curve method to show the difference in their expected performance in the field.
An example of the creep curves is shown in the figure below:
Samples A, G, H and J are TOR materials approved for use in the field, and are expected to perform well in this test. Samples F and L are “dual products” – those which can be used on both flange and TOR. Samples B, C, D, E and K are flange products, expected to have low traction.
The creep curve method was developed at Ingram Tribology between 2018 and 2021 [3]. The method has been used in a inter laboratory study (round robin) and is now included as an indicative test in the EN 15427 standard for TOR materials and flange products.
We conduct creep curve testing on TOR materials and flange products as part of our contract testing offer. This method helps the development of new materials for rail/wheel contacts. The method also acts as an accurate screener before full scale track testing.
References:
[1] Thompson, D., and Jones, C., Noise and vibration from the wheel-rail interface. Wheel-rail interface handbook. CRC Press 2009
[2] Eadie, D.T., Santoro, M., Powell, W., 2003. Local control of noise and vibration with KELTRACK friction modifier and Protector trackside application: an integrated solution, Journal of Sound and Vibration, 267, 761-72.