How to increase endurance of welded joint?

Coercimetrical monitoring as a base of high long-term endurance of the welded joint at the example of railway


It goes without saying welded joint should be defect-free. But not each sound welded joint can operate without cracks and fractures within its design life, i.e. until its geometrical parameters would change unacceptably due to metal wear by friction. Structural and phase uniformity (homogeneity) of all metal constituents of WJ: weld, heat-affected zones and the base metal should be provided. Without special heat treatment immediately after welding achieving uniformity of deformation, uniformity of strength is generally impossible. Without heat treatment (or after not optimal heat treatment) weld becomes a strain concentrator and softened heat-affected zones came to beginning of failure faster than base metal and weld:


How this can be “seen” by non-destructive method another than hardness testing? A much more efficient for this is method which use coercive force. Value of coercive force, Hc, depends from the material treatment. Coercive force of weld is much higher than of base metal and coercive force of heat-affected area is lower than base metal.


Mechanical (cyclic and static) endurance of a weld joint, WJ, just like of any mechanical structure, is ensured by uniform strength of its main parts – the weld, the heat-affected zone and the base metal. Power engineers by intuition and from operating practices have long understood the need to ensure homogeneity of the WJ in order to achieve trouble-free operation during its lifetime. WJ of annular welded joints on coolant pipelines of nuclear and heat power plants which, upon stripping the reinforcing weld, could neither visually nor coercimetrically be detected (identified) on the pipe, so indistinguishable from the main pipeline metal it was. Special Scientific Engineering ascertained that fact for its over than 30 years’ experience.  In terms of coercimetry (and in terms of hardness) such welded joint obtained with optimal technology of welding and heat treatment looks as follow (yellow line at the picture below):

Pic. 2

Optimal and real welded joint proceed in different ways during service life. In terms of coercimetry they are well-distinguished in initial state and during service life. A susceptibility to fracture of optimal WJ almost doesn’t rise.

Structural homogeneity of welded joint with optimal heat treatment in comparison with heterogeneity of welded joint without heat treatment is perfectly illustrated by frequency of coercive force, Hc, values:


The method of achievement of structural homogeneity of all WJ components in the case of power engineering was very simple exhaustive search of welding technology options resorting to bench trials and verification of operation results. Wherein power engineer tries to provide not only and not so much the maximum strength of the weld itself, but strives to ensure that both the weld and heat affected zones should resist operational loads all in the same way: just as the base metal. The same method of welding technology improvement was successfully applied on railways.

At the picture below main types of deviation of WJ metal state from optimal are represented. These deviations are results of failure in welding and heat treatment and these days are not monitored.


Only with such target-oriented welding technologies all parts of the WJ resist loadings primarily as a whole formation rather than as weld-connected heterogeneous metal bands. In such WJ stress concentrators manifest themselves to a lesser extent, they occur when some part of the WJ under the effect of loading deforms more (or less) than it’s any other part. Right at the interface between such deformationally different parts stress concentration occurs. In so doing the degree of concentration is the greater, the more deformationally different are adjacent metal domains. But when the technology is achieved by way of mechanical bench testing samples of welded joints produced under different welding and heat treatment modes with no control parameter there is no guaranty of stability and reproducibility.

Coercimetry can not only significantly cut time for obtaining of optimal welding technology (including heat treatment) but makes it accompanied by controlling of susceptibility to fracture that is even more important.

How it can be realized at practice is shown at the example of railways.

We estimated the existing welding rail joints technology applied under conditions of stationary rail welding enterprise handling 600-meter strings to be laid in tracks of high-speed trains. To do so were conducted manual coercimetric measurements on freshly welded strings. Strings have passed hand over control, and were ready to be shipped for laying in the high-speed train track. There were examined totally 100 welded joints. Welding method used was electric resistance one. Measurements of the coercive force of the metal were performed directly on the weld, on both heat-affected zones (HAZ) adjacent to the weld (±50 mm) and in the areas of base metal of rail (BM) (±300 mm), where the effect of welding on the metal properties is no longer felt. Equipment: Magnetic structurescope (coercimeter) MC-04H-2.

Summary coercimetrical analysis of continuously welded rails (each rail and each welded joint) was conducted manually on stalk truck. All continuously welded rails had outgoing control and were ready for shipping. Totally 4 continuously welded rails 600 meters each and 100 welded joints on them were tested.


Control samples of WJ are done before welding of each continuously welded rail. The samples are subject of bench tests with premeasured bending loadings toward head and toward base.

Pic.5.1                                                                        Pic. 5.2

There is a result of measurements on rail metal outside of bending loading – 10,9 A/cm at the Pic.5.1.  and a measurement at the weld in the area of maximum of bending loading – 16,2 A/cm at the Pic. 5.2.

The difference of 50% shows high informational sensibility of coercive force to loadings and deformations of rail metal.


Control sample of bending bench test these days is considered as representing the state of corresponding continuous welded rail in whole. But welded rail contains 25 welded joints and such selective check does not bring out deviation in welding of each particular WJ. 100% outgoing control of structural and phase uniformity is required and today these can be provided by coercimetry only.

An ideal deformationally homogeneous welded joint should have the values ​​of coercive force in the weld and in its any zone, equal to the coercive force of base metal. For assessment proximity of actual and ideal WJ was plotted the graph. Across the whole sample of tested joints were plotted graphs of distribution of mean arithmetic values ​​of the coercive force on both sides of the weld to within ±300 mm within the same limits a schedule arithmetic mean deviation (dispersion analog) of the coercive force value Hc. Scatter of DHc values ​​obtained was 2,9 A/cm. Such scatter, taken account of with regard for strength calculation during the remaining life, suggests that fracture of the joint can occur before the rail track in the area of ​​the joint becomes unusable due to an unacceptable change in its geometrical parameters due to metal wear by friction, i.e., within its design life. By increasing sample size of the tested joints, probability of occurrence of a joint with greater spread ​​ of Hc values within its limits will only increase.


“An average welded joint” of 50 welded joints sampling. Maximum and minimum variations.

Maximum and minimum variations are plotted on the base of coercimetrical measurements of 50 WJ at welds and heat-affected zones. The graph of arithmetic averages in each measurement point  is also plotted. There is founded a welded joint containing the maximum difference between maximum and minimum values that means that cycle strength of this welded joint is not enough for correspond to estimated service life.

The bottom graph represents minimum values. It shows the best structural and phase uniformity of all metal constituents of WJ: weld, heat-affected areas and the base metal and so proves that such optimum result is reachable by existing welding technology. But also it (in comparison with other graphs) demonstrates that the technology is not stable and so does not provide optimal result for each particular WJ. And there is no current control procedure for each WJ.

The entire analysis of welding technology performed shows that the basis for acceptance testing of welded joints should be assessment of structural homogeneity, rather than occurrence of welding defects. The welded joint is to be defect-free by definition. However, its durability can be ensured from the very beginning only based on its structural homogeneity. Only in this way it can be ensured already at the welding stage in the production of strings of jointless tracks at a stationary rail welding specialized enterprise. Coercimetric criteria of the degree of structural homogeneity of WJ are distinct and physically grounded. The control procedure of joint for this parameter can be easily fully automated – from the actual measurement process to the decision-making process (with regard for serviceableness), inclusive. On this basis, it is easy to create an electronic databank of the initial state of each joint. Subsequent operational cracks and fractures in the joints are an inevitable consequence of structural inhomogeneities in WJ.

Providing high long-term endurance of the welded joint is a multiparameter problem. Presently established welding technologies levels, as well as the scope of knowledge about the fracture mechanics, are quite high and relatively balanced.

Such multiparameter problem today does not have any pronounced weak links. All components of the multi-parameter function are already about equal in weight for ensuring WJ endurance. But this level is not high enough so that actually reached duration of smooth operation of the welded joint within the estimated service life could cease to be relevant.

The coercive force due to its physical nature is an effective integrated materials-research informational characteristic which is very sensitive to almost the entire set of processes in the weld that affect the endurance of the welded joint, from the micro-level up.

For rail manufacturer Implementation of 100% control of welded joints provides high long-term endurance of welded joints, reduces rail joint fractures, enable achievement of best welding technology for each particular kind of operation with its further repeatability.  For railways coercimetry is capable to reduce the cost of operating the rail flaw detection services while increasing reliability of the track.


On the base of the report at WCNDT 2016 “Coercimetric Technological and Acceptance Testing of Welded Joints to Ensure their Useful Life as Exemplified By Butt Welding of Rail Joints, including Subsequent Operational Diagnostics” by Gennadiy BEZLYUDKO (Special Scientific Engineering, Ltd, Kharkiv, Ukraine), Roman SOLOMAKHA (Special Scientific Engineering, Ltd, Kharkiv, Ukraine), Anastasiia LUKINA (National Technical University "KhPI", Kharkiv, Ukraine)

The original version of the report is available at




Coercimetric Technological and Acceptance Testing of welded joints accompanied with graphs.pdf - 1177.39 Кб