The new way

22 November 2005

European standards body CEN TC 147 has issued a crane standard that has a different view of crane fatigue than the previous FEM standards. Ilkka Riikonen explains how EN 13001 differs from what came before and how it will be used in future.

This new standard series do not intend to change the "level of safety" that has been built in the former European standards (national and FEM), in general. The records of using the former standards are mainly positive and satisfactory.

However, we have learnt more since the background of the former standards was established (30-40 years ago).

Rudolf Neugebauer from Darmstadt Technical University in Darmstadt, Germany was the first to voice his concerns about the standard. He first developed these concepts almost 20 years ago and presented them first in ISO/TC96/SC1. During his career he had studied reasons for numerous failure cases of cranes and come to conclusions that it was necessary to determine the real stress cycles more accurately than in the past. He started the drafting of EN 13001 as the convenor of CEN/TC147/WG2 until he retired in the early 1990s.

The final standard contains some new features that require learning among designers and even salesmen – the limit state method and a new classification system - and therefore has faced criticism. But I trust that those new features are state of art and are necessary for the crane safety.

The limit state method

The limit state method is not exactly totally different from the allowable stress method currently used, but it is a more general method that provides consistent safety margins in all parts of the structure. The allowable stress method is one special case of the limit state method obtained so that all partial load coefficients are assigned the same importance. For example, when a crane lifts a load 1.5 times its maximum capacity, the allowable stress method supposes that the balancing own weight of the crane (including a possible counterweight) is also increased by the same factor. This does not happen in nature and therefore the allowable stress method may lead to a less safe design.

The use of limit state method will not cause any significant changes to bridge crane design and to gantry cranes without cantilevers. For cranes which lift loads outside the rectangle defined by the supporting corners – such as jib cranes, davits, sheerlegs and derricks – it is essential to use the limit state method. This may require strengthening of some structural members compared with the former design made formally according to the allowable stress method. Use of the limit state method means the upgrading of some critical details to the same safety levels as the other parts of the structure.

The effect of new fatigue analysis to all types of cranes is not straightforward. According to our experience so far the cranes have a lot of details, which are not at all designed up to the permissible fatigue stresses and those points should be found by correct stress history analysis of the components. On other hand, other points may be found where the stress levels used in the past are exceeding the permissible fatigue stresses. In general, I’d say, safety and reliability will increase, and some savings in manufacturing may also be expected.

New classifications

The purpose of classification is to take into account the huge differences that exist in the intensity of use of cranes. Some cranes make a work cycle (raising and lowering a load) every minute, totalling to more than two million cycles in life with almost 100% of the rated load. Other cranes may make less than 20,000 work cycles during their whole life, and the loads may be mostly much below the maximum rated capacity. The difference between these two cases is greater than 100 times. This difference leads to quite different wear and fatigue life of components, if not adequately considered in design.

The FEM crane design standard (1.001: 1987) dealt with this variety by quantifying intensity of use into as many as ten different levels. It considered the intensity of use on three scales: on the level of the crane as a whole, at the level of component mechanisms (such as hoist units), and at the level of individual components (such as wire ropes or sheaves). Although some of this work was based on ISO 4301-1, FEM introduced classes of use for each individual component based on the stress spectrum factor and number of actual stress cycles.

EN 13001-1 has same basic classes as FEM and ISO for load spectrum factor and working cycles, but these are not combined to one class of the crane as a whole. These are the familiar A1-A8 classes. The message for the users of this standard – manufacturers and users of cranes – is that the parties shall consider and agree the real intended work cycles of the crane: number of cycles and masses of the loads. It is not enough to make an educated guess of A-class, because it hides the basic parameters of crane use and does not provide enough information for the development of real stress histories in individual components.

EN 13001-1 also totally omits the classes of mechanisms - this may be surprising. It sounds logical that more running hours and a heavier load spectrum produce more damage than less hours and light loads, and so the M-class should be used as a design parameter. However, the stress cycles in individual components do not necessarily correlate to the run time hours. In hooks the fatigue effect depends on the load spectrum and the number of lifts – not on the time used for lifting. Damage in ropes depends on the load spectrum and on the number of bends in a particular point of the rope. The number of bends depends on the rope reeving system, the height of lift and the hoist speed in addition to the time used. In gears most shafts and gear wheels shall be dimensioned for infinite life anyway. Therefore, there are not so many components in a hoist mechanism which could be calculated directly on run time basis – except perhaps some bearings in gears.

Here is an example of how the M-class system can go wrong. There are two factories which produce similar products at same rate. Factory A thinks that it can manage with a crane that has hoist speed of 4m/min and class M6. Factory B thinks that it can save time by buying a similar crane with a hoist speed of 16m/min. In factory B the crane makes the same lifting cycles with same loads as at factory A, but uses only a quarter of the time. As the run time is only one quarter of that in A’s crane, factory B can choose a class of mechanism two steps lower. Consequently, factory B will get a thinner rope and smaller drum to-rope diameter ratio (D/d). This will lead to much shorter life of the rope in the same job!

Instead of M-classes, EN 13001-1 defines new classes for average displacements of movements (of hoisting, travelling and slewing) and average number of accelerations in one work cycle (for all motions).

By omitting the classes of mechanisms, EN 13001-1 connects the design parameters of the mechanisms directly to the usage parameters of the whole crane. In prior practice the classes A, M and E might have been estimated separately and the result might have been internally contradictory.

Applying the classification system

Serial hoists, as well as individual hoist mechanisms, shall be classified according to the above-mentioned parameters of EN 13001-1, Q, U, D and P classes. Q- and U-class should be labelled in the machinery and all should be included in the documentation.

All of these classes should be agreed between the purchaser and the seller. When these parameters are fixed and the operation speeds are specified, the designer can calculate further all necessary parameters which are needed to determine the stress histories (i.e. the stress spectrum factor x number of stress cycles) for each component that is important for the design process.

The classified parameter stress history factor 's' can be used directly for the determination of the maximum stress amplitude permissible for that component in the specified use. In fact, the factor 's' did appear in the formulae of FEM 1.001, but was not so named. The stress history parameter is just a new presentation of Palmgren – Miner’s rule for cumulative damage.

The crane buyer and/or user should not define the stress history parameters or classes, as they are unlikely to have the competence to do so. The stress histories of different components require special design skills. Also, determination of the s-parameters by the buyer transfers responsibility from the designer. This violates the principles of the European Machinery Directive that the manufacturer is responsible for design.

Measured drawing for overhead crane Measured drawing for overhead crane