Comprehensive Evaluation of Properties of Laser-Welded Overlay of Powder H13 Steel on Structural S355 Steel and on H11 Tool Steel

Abstract

Laser hard overlaying is an advanced, perspective technology with wide industrial applications, for example, dies. The aim is to improve surface properties like wear resistance using special layers of powder sintered or remelted by laser beam. At present, dies are manufactured by machining with following bulk heat treatment, which is an expensive process particularly due to use of expensive high-alloyed tool steels. Repairs performed using arc or plasma welding introduce a big amount of heat to the part, which can cause dimension changes and material degradation. These methods often fail also due to low weldability of the materials. An advantage of laser overlaying is minimization of these difficulties. The paper contains a comprehensive evaluation of several types of hard overlayed powder of H13 tool steel on a S355 structural steel and on H11 tool steel using a laser beam. Macro- and microstructure, hardness and fatigue resistance are evaluated, including fatigue damage mechanisms. In the case of the H13 welds on the S355 steel plate, the quality of the welds was mostly acceptable, without pores or segregate impurities and with a good interconnection between the weld track and base material. Results are completed with basic measurement of residual stresses using destructive strain-gauge methods. High tensile residual stresses of 1465 MPa were measured at the boundary of the first track of the single-layer overlay. Fatigue resistance is sensitive on surface and subsurface defects, which can significantly reduce endurance limit. Fatigue strength of specimens with the single layer overlay was considerably lower than fatigue strength of the S355 steel. The decrease was between 25% and 50%. In the case of overlay of H13 on H11 tool steel, the decrease in fatigue strength was between 25% and 30%.

1. Introduction

Additive Manufacturing (AM) is a new innovative technology that allows the direct fabrication of complex, individual and high-strength metal products, based on their 3D data. The technology generally refers to processes where the computational design is used for the production of components through the addition of material. Contrary to conventional manufacturing methods, AM is based on production through layer increments [1]. In the AM of metals, metallic powders are totally or partially fused by energy of a laser and transformed by layers into a solid component that has practically no geometric constraints [2].

In spite of numerous advantages of AM like a significant reduction in waste, reduction in manufacture time, and reduction in labour costs, an application of AM is connected with numerous challenges. One of the main targets of AM is to manufacture parts to the net-shape or near-net-shape, which are characterised by no or minimal surface processing for targeted applications. This requires an in-depth understanding of the formation of various types of AM surfaces, including the variation in surface condition and controlling factors, and their influence on mechanical performance [3].

Fatigue resistance of numerous AM parts has to be considered, as the parts are usually dynamically loaded in service. In this area, a lot of knowledge can be found in the literature. Factors controlling the fatigue resistance are namely complex, affecting each other. Some examples can be given as follows.

An important factor is surface roughness, particularly, when no additional final machining is applied. Surface roughness is often affected by the presence of partially molten particles. Surface finishing treatments improve the surface quality and the fatigue life. However, the achievement of the lowest surface roughness does not necessarily correspond to the best fatigue performance, thus suggesting that multiple mechanisms may be active and that besides surface roughness also residual stresses contribute to increase the fatigue strength [4].

Another factor is porosity, which in the case of laser powder-bed fusion may depend on different laser velocities [5]. Inappropriate process parameters can cause either under-melting or over-melting. In the former case, lack of fusion and balling defects [6,7] can be formed, whereas in the latter case, keyhole pores [8] and spatter particles [9] are often observed. Under cyclic loading, these pores create local stress concentration and trigger fatigue crack initiation. Results from existing studies highlighted the influence of defects, such as porosity and surface roughness, as the cause for the premature fatigue failure of L-PBF parts [10,11,12,13]. AM metal parts are usually characterised by high strength, which can experience up to a 40% reduction in fatigue performance due to manufacturing defects [14]. It is well known that high strength positively affects fatigue resistance of metals with a low surface roughness and not containing defects. On the other hand, high-strength metals are sensitive to crack-like defects and surface microscopic notches, which initiate fatigue cracking, particularly in high-cycle or very-high-cycle regimes at low stress amplitudes, when the failure may occur after many cycles. A good indicator of notch sensitivity is fracture toughness [14].

The fact that high-strength metals are sensitive to crack-like defects and surface microscopic notches is supported by results published in [15]. The Inconel 718 steel produced by AM had a higher microhardness and higher tensile strength than the forged and rolled material, but its fatigue performance was lower. The steel manufactured by Laser Beam—Powder Bed Fusion (LB-PBF) demonstrated shorter fatigue life, especially under low and medium stresses, i.e., in high-cycle regime. The shorter fatigue life of the material obtained by PBF-LB was attributed to typical process defects besides microstructural differences, in connection to the higher sensitivity to defects.

Effects of post-processing treatments on fatigue performance of an alloy manufactured by LB-PBF were studied in [16]. The material was the Ti6Al4V alloy, which is known to be highly susceptible to the notch effect. The material was post-processed by machining and combinations of alternative mechanical and electrochemical surface treatments. Compressive residual stresses were introduced in all surface-treated samples. After tribofinishing, surface roughness was reduced to 0.31 ± 0.10 µm, which was found to be the most critical factor. It was concluded that machined surfaces presented fatigue behaviour comparable to wrought material. Additionally, alternative surface treatments showed a fatigue behaviour equivalent to the casting material. The results concerned the titanium alloy, but can be generalised to other metallic materials.

Another important issue concerning fatigue resistance of additively manufactured metals is the building direction. As an example, the anisotropic fatigue behaviour was shown in [17], where an additively manufactured AISI 316L Stainless Steel had higher fatigue strength for horizontally built specimens compared to the vertical building direction. This result can be generalised. Similar results were namely obtained for the 3D printed Inconel 718 Alloy—higher fatigue strength for horizontally built specimens [18].

It can be pointed out as a conclusion that fatigue resistance of additively manufactured metals is a complex phenomenon affected by many different factors, often acting mutually. In spite of the fact that computational modelling is being used (e.g., [3]), this field usually needs a lot of experimental works targeted at specific cases of AM with the aim to optimise manufacturing parameters and considering specific service conditions of AM parts and components. The aim of this work is an evaluation of properties of laser-welded overlay of powder H13 steel on structural S355 steel, which is a cost-effective solution, as the S355 steel is quite a non-expensive material. The second part of the project concerned welded overlay of powder H13 also on H11 tool steel, which is important particularly for repairs of tools and dies. The technologies are to be used for an advanced, low-cost and effective manufacture of tools with a high temperature and high wear resistance like dies, and for repairing such tools. As these tools are usually loaded also dynamically besides wear, the investigation described in this article was targeted on fatigue properties.

2. Materials and Methods

2.1. Overlaying of H13 Tool Steel on S355 Structural Steel

The base metal was a plate of commonly used S355 structural steel. On the surface of the plate of the thickness 25 mm, different configurations of layers of the H13 tool steel were welded using laser beam. For the purpose of welding, the H13 steel was in a powder state, namely powder LPW-H13-AABM, granulation 44–105 μm. The powder was produced using the technology VIGA, i.e., gas atomised. Surface speed of laser beam was 0.5 m/min. Laser overlaying was performed at MATEX PM s.r.o. company, Plzeň, Czech Republic, using robotic laser cladding workstation, consisting of KUKA KR16 (Augsburg, Germany) robot, Laserline (Laserline GmbH, 56218 Mülheim-Kärlich, Germany), diode laser LDF 5000-40 and powder cladding optics Precitec YW-52 (Precitec GmbH & Co. KG, Gaggenau, Germany).

As the first step, one layer of a single H13 surface track was welded on the plate surface—Figure 1. Note that the plates were marked by a marker pen, by hand, using a decimal comma (,) as a symbol to separate the integer part from the fractional part of a number. The use of decimal comma is typical for continental Europe (including Czechia), unlike UK, USA and numerous other countries, where decimal point is used. The second configuration was represented by the weld of five partially overlapping tracks, where the tracks were in a single layer. The third configuration was represented by five tracks like in the previous case, but in two layers, layer upon layer—Figure 1 at the bottom. Time between layers was long enough to enable complete heat dissipation, and the layers were not thermally affected by previous ones.

As regards the order of the tracks in case of five tracks overlays in Figure 1, the order is from top to bottom. That is why the bottom track is the widest, because it was welded as the last one, not partially overlapped by another track like the previous four ones.

In all three cases, laser overlaying was carried out in a protective atmosphere of argon by laser beam of the power 4.5 kW. The beam surface speed was 0.5 m/min.

In the first step of experimental works, metallographic analyses were performed on transverse cuts of the plate with the welds. Metallographic samples were prepared by grinding and polishing. Macro- and microstructure was displayed by Villela-Bain metallographic etchant. Metallographic analyses were performed on the Zeiss Axio Observer light microscope (Carl Zeiss Microscopy GmbH, Jena, Germany).

Hardness was measured using the Vickers HTM 7307 hardness tester (Grayford, Kent, UK). Hardness of welded layers and courses of hardness were measured as HV10 Vickers hardness. Two directions of HV10 courses were evaluated, namely parallel or perpendicular to the surface, respectively. The parallel course was evaluated at the depth of several tenths millimetres from the surface, in both single-layer and three-layer welds. The perpendicular course comprised both values from the welds and base material.

2.2. Overlaying of H13 Tool Steel on H11 Tool Steel

Overlaying of H13 Tool Steel on H11 Tool Steel was another research field. Unlike overlaying on the S355 base metal described above, where investigations were important for manufacture of new dies, the importance of this investigation arose from practice, when tools made of the H11 steel have to be repaired by further welding of the H13 steel. This welding technology was carried out at RAPTECH, s.r.o. company, Zruč-Senec, Czech Republic, which, however, has recently gone bankrupt. Nevertheless, the experimental project works have been successfully finished, and the bankruptcy affected neither the research project nor the results.

The overlaying of the AISI H13 powder on the base plate of AISI H11 (1.2343) tool steel was performed by diode laser Laserline of the maximum power 3 kW and Precitec laser head YW52 with four nozzles for powder feeding. The powder feeding was performed with the Model 1264 Powder Feeder with a possibility of preheating. The powder was preheated to 88 °C. Laser beam surface speed was 0.5 m/min.

As before, different configurations of tracks were welded on two base plates of the H11 steel of the thickness 25 mm, namely tracks 1–4 on the first plate and tracks 5–7 on the second plate. Tracks 3 and 6 were welded with the same parameters, representing three multiple layers, layer upon layer, of five neighbouring tracks partially overlaying each other. Track 2 represented a single layer of five neighbouring tracks partially overlaying each other. In all cases, the overlapping distance was 1.7 mm, whilst the laser beam diameter was 2.8 mm.

Specimens of tracks 3 and 6 were examined from the point of view of macrostructure on a metallographic optical microscope. Two specimens of track 3 and one specimen of track 2 were tested by fatigue.

2.3. Evaluation of Residual Stresses

Another field of research was targeted at evaluation of residual stresses caused by the welded layer. Residual stresses affect fatigue life and endurance limit particularly when different types of defects are present, e.g., surface microscopic notches, pores, microcracks [4,16]. That is why residual stresses also were evaluated in this work. The measurement was carried out using a destructive method, with strain gauges. Residual stresses perpendicular to the weld tracks were evaluated. Two strain gauge (SG) chains, Hottinger Baldwin Messtechnik HBM 1-KY11-4/120, the largest HBM chains, were used. The mutual distance of the SGs in the chain was 4 mm, as schematically shown in Figure 2.

Placement of the strain gauge chains is shown in Figure 3. The first chain with SGs 1–10 was glued across the five double-layer overlay, in the bottom part of Figure 3. Strain gauges were attached to completely cooled specimens. A two-component composite HBM adhesive was applied. Unlike a simple high-speed adhesive, this adhesive is particularly suitable for wavy surfaces. The first two strain gauges were outside the overlay, the third SG on the boundary between the widest weld track and base metal plate, the ninths SG on the opposite boundary and the tenth SG in the gap between the two five tracks overlays. As regards the second chain glued across the single-layer overlay, which was slightly narrower, the second SG was on the boundary, the seventh one on the gap between the overlay and the single weld on the top of Figure 3, SGs 8 and 9 on the boundaries of the weld. SG 10 was near the margin of the single trace, on the base metal plate. Measured strains were evaluated using strain gauge device Hottinger Baldwin Messtechnik UPM60, which has an excellent temperature compensation and stability. The device was computer controlled and recorded.

The measurement was performed by sequential cutting of the welded sample and final removal of the base metal (BM) by turning. Hole drilling methods were not applied. The turning was performed with a specific clamping of the plate, when the rotation axis was perpendicular to the plate surface. It was not technically easy to do it, but it was eventually successful. The principle of the method of the residual stresses measurement consisted in the fact that before starting the cutting and material removal, the system of the base steel plate with the welds was balanced. The sequential cutting gradually changed this balance and after the final removal of all base metal by machining; the residual stresses relaxed and measured strains corresponded to absolute value of the stresses, with the opposite sign. Stresses were calculated from the strains using the E-modulus E = 206 GPa.

The cutting and material removal was carried out with the following steps:

−

Initial strain measurement after setting the strain gauge device to zero;

−

Cutting the bottom layer of BM to the thickness approximately 4 mm: just the BM plate excluding the overlay thickness;

−

Further cutting the bottom layer of BM to the thickness approximately 2 mm excluding the overlay thickness;

−

Cutting of the specimen to three separate pieces—the part with double weld, the part with the single-layer weld and the part with the single weld track; the strain gauge chain was cut, too;

−

Final machining all remaining BM off bellow all the welds, as indicated in the Section 3.2.

At each step of cutting or machining the base metal off, always the whole base plate was machined, not only material under the tracks. After the final machining, all residual stresses were fully relaxed and original residual stresses could be simply calculated from the measured strains. Note that the advantage of the used method is that it is integral and is able to evaluate residual stresses into the depth, unlike, e.g., X-ray method being able to measure just very thin surface areas. It also should be pointed out that the machining was performed carefully, very slowly, i.e., so-called low-deformation machining so as not to introduce any additional residual stresses. The temperature was never increased.

2.4. Fatigue Tests

Evaluation of fatigue properties was performed because of the need for knowledge about resistance of welded layers to fatigue loading, as cyclic loading is an important part of the service loading in such situations, besides thermal and wear loading.

Fatigue tests were performed on small specimens of approximately square cross-sections of 7 × 7 mm or 9 × 9 mm under three-point bending—Figure 4. Dimensions of specimens were exactly measured individually for each specimen. Test span was 35 mm. The central point of force was at the side opposite to the weld surface, which was therefore loaded by tensile stress with the maximum value at the centre decreasing to zero at the marginal points. In Figure 4, both double-layer specimens (left) and single-layer specimens (right) are shown. The supports and central point of load are indicated by red arrows. The specimens in Figure 4 are after fatigue cracking, which will be discussed in the Section 3.

Fatigue tests were performed on a SCHENCK PHG high-cycle fatigue machine, manufacturer Carl Schenck AG, Darmstadt, Germany. The test frequency was 40 Hz. The SCHENCK PHG, often referenced in the literature as a SCHENCK-type pulsator or PHG 3000 N, is a high-frequency fatigue testing machine used for determining the fatigue strength and life of materials. It works on mechanical cantilever principles, not servohydraulic, with frequencies up to 50 Hz. R-ratio was 0.1, the minimum possible to secure the specimen against falling out. Target number of cycles for evaluation of the endurance limit was 10 million. In the case of overlaying of the H13 steel on the S355 base metal plate, the additively welded surface layers were in the as-built state, with no machining or grinding. In the case of overlaying of the H13 steel on the H11 tool steel, a thin surface layer of each specimen was ground off after careful measurement. The ground layer is indicated in Figure 5. The specimen axis was always perpendicular to the welding direction.

3. Results and Discussion

3.1. Optical Microscopy

Macrostructure and microstructure of the welds and base metal were analysed. In the case of the H13 welds on the S355 steel plate, the quality of the welds was mostly acceptable, without pores or segregate impurities and with a good interconnection between the weld track and base material—Figure 6. Microstructure of the base material was, as supposed, ferritic-pearlitic, with uniform grain size—Figure 7. The heat-affected zone was approximately 0.9 mm wide, refined in the direction to fusion zone, which was formed by martensitic structure. The width of the heat-affected zone was measured on the first track of a double layer weld as the distance between the deepest point of the weld and the end of the heat-affected microstructure. The welds were formed by martensitic structure of a dendritic type—Figure 8. Both martensitic and dendritic structures were analysed by optical microscopy only; neither scanning electron microscopy nor phase identification techniques were applied.

From the macroscopic point of view, the outer weld track was problematic, which contained pores and lacks fusion—Figure 9. According to the literature, the defects may depend on different laser parameters including velocities [5].

Total views of macrostructure of the single-weld and the single-layer overlay with five tracks after etching are shown in Figure 10 and Figure 11, respectively. The laser tracks (beads) and heat-affected zone with a thickness 0.9 mm are well visible. The lacks of fusion at track edges, which affected fatigue resistance and fatigue crack initiation, will be discussed later.

3.2. Evaluation of Hardness

The method of evaluation of HV10 hardness was described in Section 2.1.

The course of hardness perpendicular to the surface, i.e., in the direction from the surface through laser tracks to base material, is shown in Figure 12.

In Figure 12, maximum tensile strength values Rm are estimated according to [19]; the maximum and minimum hardness, respectively, are shown, too. Note that the lowest strength, 440 MPa, which corresponds to the base metal plate, is quite low, outside the range typical for the S355 steel 470–630 MPa [20].

The courses of hardness in the direction parallel to the surface in both the single-layer track and double-layer tracks are in Figure 13. It follows from the diagram that hardness values inside the overlaying tracks were always strongly affected by the next neighbouring track, causing a drop in hardness. Note that in Figure 13, tracks were welded from the right to the left, so the leftmost points in the diagram correspond to the last track, no more affected by further welding. The experimental points approximately correspond to the macrostructure photograph inserted into the diagram.

3.3. Analysis of Residual Stresses

The method of measurement of residual stresses is described in Section 2.3. The effect of gradual machining on measured strains in double overlay is shown in Figure 14.

The shape of the residual stresses curves in Figure 15 are logically consistent with the curves in Figure 14; they just have an opposite sign. Also note that when the machining of the base material was just partial, to the remaining thickness of the BM plate of approximately 4 mm excluding the overlay thickness, the residual stresses on the weld surface were slightly negative, giving evidence of a complex, complicated integral stresses distribution and balance in the weld and below it. Residual stresses reached into the base plate considerably and were balanced there. After the final machining, where the base metal was completely removed off, as indicated by the red dashed line in the macrostructure photograph in Figure 15, original residual stresses in the weld layer completely relaxed, resulting in the biggest negative measured strains particularly at the boundaries between the overlay and the base metal. The corresponding positive peaks of the recalculated residual stresses were quite high, around 800 MPa, and were located at the marginal boundaries of the overlay.

What is most crucial are the very big tensile residual stresses at the boundaries of the overlays, affecting fatigue life and failure mechanisms. Inside the overlays, residual stresses were quite moderate and fatigue cracks therefore did not occur there, just at the boundaries between the overlays and base metal.

Quite an important result is shown in Figure 16, namely residual stresses in the single-layer overlay. The peak at the boundary of the first track was considerably higher than in the case of the double-layer weld, namely 1465 MPa. The peak on the opposite boundary was in contrast lower, only 370 MPa. The more uniform distribution of residual stresses in the double-layer weld can be explained by the fact that during overlaying of the second weld, the original first layer was tempered, resulting in reduction in the stresses. The comparison of residual stresses in the single- and double-layer overlays is shown in Figure 17. Note that the biggest value of residual stresses is quite close to the maximum strength of the weld evaluated from hardness, which was 1950 MPa. Such big residual stresses can considerably affect mechanical and particularly fatigue properties.

3.4. High-Cycle Fatigue Resistance and Fatigue Cracking Mechanisms

Results of high-cycle fatigue tests are shown in Figure 18. Stress range was calculated considering not only the nominal dimensions, but calculations were corrected and recalculated considering the actual specimen height at the cracking point including the overlay.

Concerning the scatter, the laboratory usually performs statistical evaluations and regression analyses, usually with confidence bands and tolerance limits. However, in the case of limited and very limited number of specimens, it does not make sense. On the other hand, the differences in the S-N diagram can be considered as unequivocal. Particularly interesting would be to analyse specific reasons of exceptionally lower fatigue life of the two specimens in the diagram, namely investigating fatigue crack initiation mechanisms. Unfortunately, it was no more possible within the project. There is a field for further research.

Fatigue strength of base material S355 steel was high. Fatigue strength of overlaid specimens was considerably lower, particularly in the case of single-layer overlay. An interesting phenomenon is that fatigue life at the fairly high stress range around 485 MPa is practically the same for almost all tested batches—base metal, double layer, single layer and also H13 weld on H11 tool steel. This indicates that fatigue life was affected by the properties of overlays, by resistance to fatigue crack initiation and early growth, rather than by the base metal. Number of cycles to failure at 485 MPa stress range was around 50,000. Considering the thickness of overlay approximately 2 mm, then the corresponding mean fatigue crack growth rate would be 4 × 10⁻⁸ m/cycle, which is a reasonable value, typical for region of stable crack growth [21]. Nevertheless, it would be interesting to evaluate fatigue crack growth (FCG) resistance of overlays mutually compared and compared to FCG of base metal. Unfortunately, such measurements were not possible within the performed research project carried.

In Figure 18, there are two points of premature failure: one concerns the double layer, one the single layer. The premature cracking probably occurred due to surface defects. Some of them, namely not completely welded powder particles, are documented in Figure 19. Another phenomenon worthy to discuss is the point with exceptionally high fatigue strength of the H13 double-layer overlay on the H11 steel, namely stress range 598 MPa and 42,000 cycles to failure. The point is indicated by circle. The likely explanation is a lower occurrence of pores. In the single-layer overlay on H11, there were probably less pores than in the multiple-layer overlay, where pores were frequent. The fatigue strength is much higher than the extrapolated S-N curve of the single-layer overlay. In the single-layer overlay, penetrating the weld layer by fatigue crack results in an immediate specimen failure due to the yielding of the rest of the specimen—the S355 material. Note that the weld thickness of the single layer is around 2 mm in comparison with 3.5 mm thickness of the double-layer overlay—Figure 12. So, there is a considerably stronger support in the double-layer overlay enabling a longer FCG phase. This hypothesis is supported by the rather surprising fact that experimental points of fatigue life of H13 steel do not depend on the base metal, either the S355 steel or the H11 tool steel. The points lie on the identical fatigue S-N curve. In addition, the fatigue life did not depend on the overlay surface state. As already mentioned, unlike overlays of the H13 steel on the S355 steel plate with no surface grinding, thin surface layers were ground off in the case of overlaying the H13 steel on the H11 steel. Fatigue lives were identical regardless of whether fatigue cracks initiated on the surface defects in the former case like in Figure 19 or subsurface defects in the later case. It can be therefore concluded that fatigue life was determined by the resistance to short fatigue crack growth or notch sensitivity.

Considering the residual stresses, their role is important particularly in case of FCG from surface notches and defects, where they reduce effects of crack closure. In the case of smooth-surface and defect-free material, their role is fairly minor as they affect mean stress, not stress range. An example is the right specimen in Figure 4 and Figure 20, where cracking occurred at the notch at the overlay boundary with the residual stress peak at 1465 MPa—Figure 17.

The last, but not least comment concerns defects, namely pores inside the three-layer overlay of the H13 on H11 tool steel, when the process parameters were not yet optimised—Figure 21. Note that this photograph is “bottom-up”: the base metal is on the top. In this case, resistance of fatigue short cracks to growth was a more critical factor than fatigue resistance of smooth defect-free specimens. This also explains why the fatigue life of H13 welds on H11 steel plate was equivalent to double-layer H13 weld on the S355 plate, where fatigue cracks were initiated on the weld surface defects like in Figure 19.

The most frequent failure mechanism is shown in Figure 22. Fatigue crack initiation on surface defects, namely not complete fusion of powder particles, is followed by growth through the overlay. In Figure 22, the overlay is of the single-layer type, of the thickness slightly bellow 2 mm. The fatigue life was determined by FCG through the overlays. That is why specimens with the double-layer overlay of the thickness cca. 3.5 mm had higher fatigue life than the single-layer specimens.

4. Conclusions

The aim of this work was an evaluation of properties of laser-welded overlay of powder H13 steel on structural S355 steel, which is a cost-effective solution, as the S355 steel is quite a non-expensive material. The second part of the project concerned welded overlay of powder H13 on H11 tool steel with the potential to be used for repairing of H11 tools, e.g., dies. The results of this work are essential particularly for industrial applications of the technologies, e.g., dies, besides an improvement of scientific knowledge in the field. The main conclusions can be summarised as follows:

In the case of the H13 welds on the S355 steel plate, the quality of the welds was mostly acceptable, without pores or segregate impurities and with a good interconnection between the weld track and base material. The microstructure of the base material was ferritic-pearlitic, with uniform grain size. The heat-affected zone was approximately 0.9 mm wide, refined in the direction of the fusion zone, which was formed by martensitic structure. The welds were formed by the martensitic structure of a dendritic type. Problematic was the outer weld track, which contained pores and lacked fusion.

Tensile strength values, Rm, of overlays estimated from hardness corresponded to 1950 MPa in comparison with the 440 MPa strength of the S355 base metal. The overlay thickness was 2 mm and 3.5 mm in the case of the single-layer and double-layer overlays, respectively.

High tensile residual stresses were measured at the boundary of the first track of the single-layer overlay, namely 1465 MPa. The peak on the opposite boundary was considerably lower, only 370 MPa. Residual stresses in the double-layer overlay were more uniform, between 700 and 860 MPa.

The fatigue strength of specimens with the single-layer overlay was considerably lower than fatigue strength of the S355 steel. The decrease was between 25% and 50% of the strength of the double-layer overlay on S355 steel, similar to that of the overlay on H11 steel. The decrease compared with the S355 steel was only between 25% and 30%. At high-stress range, namely 480 MPa, there were almost no differences between all tested modifications.

Fatigue resistance of all overlaid specimens was affected by notch sensitivity and resistance of short cracks to growth from defects. In spite of considerable new knowledge generated within the work, the findings open occasions for further investigations in the field. Consequential research is being continued to further optimise the laser overlaying parameters and to minimise defects. Results are promising.