**3. Results**

We collected maintenance frequencies for different technical boundary conditions from various infrastructure managers, mainly from Europe facing comparable mixed traffic networks. In this results section, we depict the main effects of changing parameter values on the maintenance frequencies and service lives. Concluding, we also calculated the average maintenance needs for an artificial network to underline these differences.

#### **3.1 Ballast-related maintenance and renewal**

As mentioned in Chapter 2, ballast maintenance frequency and service life of track cannot be analyzed separately, at least for tracks on concrete sleepers.

We start with the influence of track loading depicted in gross-tons. Note again that tonnage is only a rough estimate for the actual loading of track, but in this case, we compare mixed traffic tracks only so that the influence of different traffic segments and vehicles smooth out or average over the network. This is true for traffic volumes up to some 25 million gross-tons. Higher tonnages in mixed traffic can only be operated with an increasing amount of freight trains. This effect cannot be depicted by tonnage only [23]. If we look at the main maintenance action for the ballast, leveling-lining-tamping, we can see an almost linear increase in tamping needs, while service life drops also more or less linearly (**Figure 4**).

The curvature defines both tamping demand and service life to a certain degree, especially if curves get narrow. Below some 600 m, tamping demand increases slightly and service life starts to drop. In very narrow curves, about one-third of the service life is lost even though tamping is more than doubled (**Figure 5**). The high cant in combination with increasing lateral forces leads to a fast loss of track geometry.

The influence of substructure quality is dominant as **Figure 6** shows: poor dewatering of the track leads to a one-third higher tamping demand, and poor soil (and proper drainage) to a doubled demand. With this increased ballast maintenance at least, the service life can be assured. If poor soil condition and poor drainage condition meet, service life can drop to half even though tamping is frequent (every

**Figure 4.** *Leveling-lining-tamping over loading.*

**Figure 5.** *Leveling-lining-tamping over track radius.*

## *Assessing Average Maintenance Frequencies and Service Lives of Railway Tracks:… DOI: http://dx.doi.org/10.5772/intechopen.110488*

**Figure 6.** *Leveling-lining-tamping over subsoil/drainage quality.*

third year) and ballast is to be cleaned one time. These technical consequences lead to the high economic efficiency of subsoil rehabilitation in such cases as the additional costs in track renewal are compensated by the saving in maintenance over the (prolonged) life cycle.

In ballasted track, of course, ballast quality is crucial. This is true for the composition of the ballast in terms of ballast stone size distribution (sieve curve) as well as the shaping of the single stones. The main trigger is the material quality though. This quality is assessed using the LA values usually (see Chapter 2). In case of magmatic material (e.g., basalt) with LA values below 14, the ballast is not the limiting component in track with loadings below 15 million gross-tons (mixed traffic) as also tamping demand is very low (every 10 years). Metamorphic material (LA values around 17, e.g., granite) leads to an acceptable tamping frequency (every 6 years for 10–15 mio. Gross-tons, **Figure 7**) and a reasonable service life. Poor ballast material (e.g., limestone) leads to doubled tamping needs on the one hand, while the halved service life compared to good ballast quality can only be achieved in cleaning the ballast once in the lifetime of track.

Moreover, the sleeper type in use triggers the tamping demand. Here, concrete and wooden sleeper perform similarly concerning track geometry stability (the service life of the wooden sleeper track is limited due to the limited average lifespan of wooden

**Figure 7.** *Leveling-lining-tamping over ballast quality.*

**Figure 8.** *Leveling-lining-tamping over sleeper type.*

sleepers, **Figure 8**). A remarkable reduction of tamping needs is recorded using concrete sleepers with under sleeper pads (USP): for the traffic load used in the example in **Figure 8**, the tamping frequency drops from every 6 years to every 12 years. As the sleeper-ballast interface provides much more contact area, stresses are lower and the ballast is protected especially at the uppermost layer, leading to a significant increase in service life.

#### **3.2 Rail-related maintenance**

This subchapter depicts the maintenance necessary for rails. Accumulated tonnage may lead to a necessary exchange of the rails due to an overridden fatigue limit. This limit is about 280 mio. Gross-tons for 49 kg rails, around 500 mio. Gross-tons for 54 kg rails, and beyond 1000 mio. Tons for profiles with more than 60 kg per meter. For the rail profile 60E1 or UIC60 which is widely used in European mixed traffic railway networks, a rail exchange due to this limit is not necessary generally looking at the achievable service lives (**Figures 4**–**8**) of 50 years in maximum. Even if the loading reaches some 100,000 gross-tons, which is somehow the maximum possible traffic volume in mixed traffic, rail replacement as a result of exceeding the fatigue limit will not be necessary.

What keeps is maintenance linked to rail surface failures and to the side wear in the outer rails of curves where the contact between wheel flange and rail head leads to heavy wear. While the latter phenomenon can only be handled by changing the rails, surface failures can be treated by rail grinding or milling as long as executed early enough. In sharp curves, corrugation waves and re-profiling are the dominant aspects, while in wider curves rolling contact fatigue (RCF) damage in form of head checks is the trigger for the rail surface maintenance. **Figure 9** shows the number of grinding interventions for different track radii.

Rail grinding cannot be handed without looking in parallel to the rail exchange. **Figure 10** gives the amount of outer rail exchange for different track radii. In brackets, we added the exchange of the inner rail which is a consequence of rail grinding due to corrugation waves on the one hand and on the other one of achieving matching rail profiles for outer and inner rail supporting smooth track guidance.

Both rail grinding and rail exchange can be reduced significantly by using higher rail steel grades. **Figures 11** and **12** show the rail maintenance frequency over the trach
