Dipl-Ing Gregor Girsch
voestalpine Schienen GmbH, Technical Customer Services
Dipl-Ing René Heyder
Deutsche Bahn AG, DB System Technology, Materials Engineering & Failure Analysis
Dipl-Ing Nicole Kumpfmüller
Deutsche Bahn AG, DB System Technology, Track Dynamics & Assessment
Rupert Belz
Deutsche Bahn AG, DB System Technology, Permanent Way Technology
INCREASING the resistance to both wear and rolling contact fatigue is of equal importance when developing new rail steels. The cost of dealing with RCF defects is very significant, both in terms of maintenance and in terms of the service life of track components, so the need to find ways to reduce these costs is increasingly important.
The experience of many railways shows that harder rail grades exhibit better resistance to wear and corrugation than softer grades1. Theoretical considerations also attest to better resistance to RCF for harder rail steels2. This was why voestalpine Schienen GmbH set out a few years ago to develop new types of rail steel: pearlitic, reaching a tensile strength above 1250MPa (head special hardened), and bainitic (lower bainite, heat-treated) with tensile strength above 1400MPa (Table I). These steels have been tested under different track conditions across the world, and results to date show significantly better resistance against wear, corrugation and RCF for high-strength head-hardened rails.
DB and voestalpine have investigated the rate of rail surface deterioration in various joint projects to determine the life-cycle costs (LCC) of different types of rail - reducing the LCC of components and products is a strategic aim of both partners3. This requires the opportunities for cost reduction in the life-cycle of track components to be identified and the potential for savings quantified. For the rail producer it is clearly important to give the customer added value by providing a product that is cost-effective.
Track tests
DB and voestalpine Schienen carried out extensive track tests with standard carbon R260 and head-hardened R350HT rail under different conditions. Examination of the damage behaviour of the two rail grades over the last five years revealed that the R350HT grade has significantly better resistance against wear. Remarkably, the tests also demonstrated that resistance to crack formation for the R350HT rail was three times higher than for standard carbon rail4,5.
Complementing the track tests, grinding trials were carried out to evaluate the number of passes required and the amount of metal removal needed to eliminate the head checks completely and to produce the target rail profile. The results showed that the amount of metal that had to be removed for head-hardened rails was much less than for standard carbon rails; this was because rolling contact fatigue developed at a slower rate.
LCC analysis for the two types of rail is based on measurements and technical assessment of the wear and crack propagation rates obtained from the various track tests. This data allowed calculation of the maintenance intervals for rail grinding which were fed into the LCC model developed by DB-System Technology. This model considers all parameters that influence LCC. Finally, appropriate grinding strategies for the different rail grades were defined1,7 in order to quantify the savings potential from using R350HT grade rail in preference to R260.
Technical assessment
An initial LCC analysis was performed for a test site at Mering on the mixed traffic line between München and Augsburg where trains run at up to 200 km/h; the track concerned carries about 90000 gross tonnes a day. The site is on a 3300m radius curve and the R350HT rail was compared with an R260 reference rail. Both are of 54E3 profile.
The general approach in the technical assessment is to calculate maintenance intervals for rail machining and the service life of rails through the wear and crack propagation rates for both grades. This data was then fed into the LCC model.
The figures for wear and crack propagation were as follows:
R260 R350HT
Wear mm 0·50 0·25
Crack depth mm 1·50 0·50
After 100 million gross tonnes, the wear for R350HT was half that of R260. This was because the site is a typical 'head check track' - RCF is the main problem at this location. The depth of rail damage caused by cracks is defined by the vertical distance from the surface of the rail to the tip of the crack. With R350HT this distance is about one-third of that for R260 (Fig 1).
The grinding tests on this section demonstrated that only half as many passes were needed to remove completely the head checks from the R350HT rail and to produce the requested nominal profile compared with the R260 grade. The amount of metal to be removed is determined by the damage depth of the cracks, the idea being that all cracks are removed.
Normal traffic also causes wear between the rail machining cycles. The maximum permitted horizontal wear limit for rail on this section was 8·4mm8. The technical service life is defined as the time the rail remains in service until this limit is reached, after which the rails must be exchanged. Hence the wear reservoir of the rail head is consumed by natural wear and by metal removal during grinding. From these data, rail maintenance cycles and rail service life were calculated for the two types of rail. Table II summarises the results of the technical assessment.
Specifying the LCC model
LCC analysis is used to aid decision-making through economic assessment and comparison of alternative strategies. Decisions in the early stages of a product's life are very important because comparatively small costs early on may incur a major share of the cost during the life of a component. It is therefore essential to consider all costs likely to occur during the entire life cycle. In many cases the analyst has to use estimated data, and the preferred method in this case is calculation of the Net Present Value for defined parameters and alternatives.
To assess the cost-effectiveness of the two different rail grades, the total annual LCC for the reference track which conforms to DB's P230 standard6,9 was investigated. The assessment provided details of the grinding cycle for the different rail grades, and it can be seen that the interval between rail renewal rises by a factor of three. So it is possible to link the renewal intervals with the costs and in this way to calculate the LCC.
Assumptions included an annual inflation rate of 2% and a discount rate of 8%, and the use of good quality ballast and subsoil. To represent the rail in the model, a product tree (PBS) can be constructed to assess the influence that the various components exert on the cost. As all track components were assumed to be identical apart from the two types of rail, the PBS can be limited to the rail.
A cost breakdown structure (CBS) is also needed, with each cost element broken down as a subset and described through equations. The costs used in the LCC analysis can be summarised as:
- Initial investment: purchase cost, including transport and installation;
- Renewal investment: purchase cost of replacement rail, including transport costs and installation;
- Disposal costs: purchase costs reduced owing to the linear depreciation at the end of the study with negative values for each component - the depreciable life of the rail as shown in Table II;
- Service life of rail: as in Table II;
- Rail grinding: as in Table II;
- Non-availability: costs are incurred with planned possessions for maintenance, as well as with unscheduled running.
Results and outlook
Fig 2 shows the calculated entire life-cycle costs for the two types of rail with assumptions built in for inflation and the discount rate of 8% according to the CBS elements for both grades. The Net Present Value is the summary of all the cost elements in the CBS.
The analyses have been carried out to show what happens if the track is not available for traffic. It clearly demonstrates that non-availability introduces heavy cost penalties, increasing costs by 10 to 15%. The graphs illustrate the importance of a well-considered investment. For example, the higher initial costs for head-hardened R350HT rail can be amortised because of longer maintenance intervals, so reducing the LCC by about 35%.
The first results for curves up to 3300m on heavily-used lines clearly show that using head-hardened rails is economically beneficial compared with standard carbon grades. The next step is to undertake further LCC analysis for other categories of track to give railway managers and engineers more information on which to base investment and maintenance decisions.
The examples used here are, of course, based on the technical standards applied in Germany. n
- CAPTION: Fig 1. The depth and angle to the railhead of cracks on the two types of rail after 85 million gross tonnes can be seen on micrographs of longitudinal sections of rail from Mering on the München - Augsburg line
- CAPTION: Fig 2. Comparison of the Net Present Value of R260 and R350HT rail, expressed in euros per metre of track, assuming 90000 gross tonnes a day, without the cost of non-availability (a) and with this cost included (b)
TABLE: Table I Advanced rail grades
Rail grade Tensile strength A5 Hardness MPa % BHN
R350HT >1175 >9 ??350
R350LHT >1175 >9 ??350
370LHT >1175 >9 ??370
UHC400 >1240 >9 ??400
DOBAIN430 ~1400 >11 ~430
TABLE: Table II. Technical assessment of maintenance intervals for test track carrying 90000 tonnes/day
R260 R350HT
Variation Grinding cycle Total load Metal Material lost Technical Metal Material lost Technical
removal through wear service life removal through wear service life
years MGT mm mm years mm mm years
1 0·5 17 0·25 0·08 12·8 0·08 0·04 34
2 1 33 0·50 0·17 12·8 0·17 0·08 34
3 1·5 50 0·75 0·25 12·8 0·25 0·13 34
4 2 67 1·00 0·33 12·8 0·33 0·17 34
5 2·5 83 1·25 0·42 12·8 0·42 0·21 34
6 3 100 1·50 0.50 12·8 0·50 0·25 34
TABLE: Table II. Technical assessment of maintenance intervals for test track carrying 90000 tonnes/day
R260 R350HT
Variation Grinding cycle Total load Metal Material lost Technical Metal Material lost Technical
removal through wear service life removal through wear service life
years MGT mm mm years mm mm years
1 0·5 17 0·25 0·08 12·8 0·08 0·04 34
2 1 33 0·50 0·17 12·8 0·17 0·08 34
3 1·5 50 0·75 0·25 12·8 0·25 0·13 34
4 2 67 1·00 0·33 12·8 0·33 0·17 34
5 2·5 83 1·25 0·42 12·8 0·42 0·21 34
6 3 100 1·50 0.50 12·8 0·50 0·25 34
References
1. Girsch G, Frank N and Pointner P. New rail grades - a technical performance overview. Proceedings of 8th IHHA conference, Rio de Janeiro, June 2005.
2. Bower A and Johnson R. Shakedown, Residual Stress and Plastic Flow in Repeated Wheel-Rail Contact. Rail Quality and Maintenance for Modern Railway Operation, Kluwer Ac Publications, Dordrecht, 1993, pp239-49
3. Garber S. Wirtschaftlichen Erfolg der Bahn absichern durch LC Logik im Beschaffungsmarkt. Eisenbahningenieur p6, January 2005.
4. Girsch G and Heyder R: Head-hardened rail put to the test. Railway Gazette International, pp42-44, January 2004.
5. Heyder R and Girsch G. Testing of HSH-rails in High Speed Tracks to Minimise Rail Damage. Wear 258, pp1014-21 July/August 2005.
6. Kumpfmüller N and Ripke B. Internal projects. Deutsche Bahn AG, DB System Technology, TZF62 Track Dynamics & Assessment.
7. Sch?€?ch W and Heyder R. Rail Surface Grinding: Exploring the Interaction. 6th International Conference on Contact Mechanics and Wear of Rail/Wheel Systems (CM2003), G?€?teborg, Sweden, June 2003.
8. Deutsche Bahn AG standard RiL 820.
9. Deutsche Bahn AG standard RiL 413.
Comparing the life-cycle costs of standard and head-hardened rail
Investigations by DB AG and voestalpine Schienen GmbH have demonstrated in the last five years that head-hardened R350HT rail has better resistance against wear and rolling contact fatigue than standard carbon rail. Data from further tests carried out to determine wear and crack propagation rates have now been used to calculate maintenance cycles and rail service life. This information has been fed into a life-cycle cost model developed by DB System Technology, and initial results indicate that it is economically advantageous to use head-hardened rail in curves of up to 3300 m radius on medium speed lines
Comparer les coûts, sur la durée de vie, d'un rail standard et d'un rail à champignon renforcé
Des investigations menées par DB AG et voestalpine Schienen GmbH ont démontré que, durant les cinq dernières années, le rail à champignon renforcé R350HT offre une meilleure résistance à l'usure et à la fatigue au point de contact du roulement que le rail standard au carbone. Des données émanant de tests ultérieurs afin de déterminer les niveaux d'usure et des propagation des fissures sont désormais utilisées pour calculer les cycles de maintenance et la durée de service du rail. Cette information a été entrée dans une modélisation déterminant le coût sur la durée de vie du rail, développé par DB-System Technology, et les premiers résultats montrent qu'il est économiquement avantageux d'utiliser des rails à champignon renforcé dans les courbes jusqu'à 3300m de rayon sur les lignes à trafics lourds
Vergleich der Lebenszykluskosten von normalen und kopfgehärteten Schienen
Von DB AG und voestalpine Schienen GmbH durchgeführte Untersuchungen haben über die letzten 5 Jahre hinweg gezeigt, dass kopfgehärtete R350HT Schienen einen besseren Widerstand gegen Verschleiss und Rollkontaktermüdung aufweisen, als aus normalem Schienenstahl bestehende Schienen. Daten aus weiteren Untersuchungen zur Bestimmung von Verschleiss- und Rissausbreitungsgeschwindigkeiten wurden nun zur Berechnung von Instandsetzungsintervallen und technischer Lebensdauer benutzt. Diese Informationen wurden in das von DB-Systemtechnik entwickelte Lebenszykluskostenmodell eingegeben, und erste Resultate zeigen, dass es wirtschaftlicher ist, kopfgehärtete Schienen in Kurven bis zu 3300m Radius auf hochbelasteten Strecken einzusetzen
Comparación de los costes del ciclo de vida del carril est? ndar y de cabeza endurecida
Las investigaciones de DB AG y voestalpine Schienen GmbH han demostrado en los últimos cinco años que el carril de cabeza endurecida R350HT presenta mayor resistencia al desgaste y a la fatiga del contacto de rodadura que el carril est? ndar al carbono. Los datos de pruebas adicionales llevadas a cabo para determinar el desgaste y el ritmo de propragación de grietas se han utilizado ahora para calcular los ciclos de mantenimiento y la vida útil del carril. Esta información se ha introducido en un modelo de c? lculo de costes del ciclo de vida, desarrollado por DB-System Technology, y los resultados iniciales indican que es ventajoso económicamente utilizar carril de cabeza endurecida en curvas de un radio m? ximo de 3300m para las líneas con tr? fico intensivo
