Revision as of 11:52, 4 July 2022 (view source) Lourdes (talk | contribs) ← Older edit	Latest revision as of 19:58, 13 April 2026 (view source) Lourdes (talk | contribs) m (Lourdes moved page Draft Blanco Salgado 404081972 to Blanco Salgado et al 2022a)
(One intermediate revision by the same user not shown)
(No difference)

---

Latest revision as of 19:58, 13 April 2026

1 Title, abstract and keywords

REPAIRING OF STEEL STRUCTURES BY COMPOSITE PATCHES ACTING AS A CRACK ARRESTORS  

REPARACIÓN DE ESTRUCTURAS DE ACERO MEDIANTE PARCHES DE MATERIAL COMPUESTO QUE ACTÚAN DETENIENDO EL CRECIMIENTO DE GRIETAS

1.       INTRODUCTION

The vast majority of ships currently in service are manufactured using various grades of steel and welding technology to produce structures capable of operating in one of the world’s most aggressive environments. Due to the effects of weather conditions, sea loads, cargo loads and operations, these structures at times suffer damages such as cracks (or fractures), corrosion or buckling which require to be repaired in an efficient and timely manner. It is standard practice for such vessels to be repaired using similar techniques to those used during the new construction phase, primarily by replacing damaged steel for new steel parts or adding steel plates by welding techniques to reinforce the damaged areas. This current common method of repair includes hot work which may require the vessel to make a ‘technical stop’ often involving gas-freeing procedures as well as undertaking many other safety procedures related to such operations on-board which in turn may result in increased operational downtime and loss of revenue for ship owners.

Composite patch repair is a comparatively easy-to-apply, robust and cheap method to extend the lifetime of many applications, within and beyond the maritime sector. Methods to strengthen and improve critical joints are already in use in other transport modes, but are new for the maritime sector, partially owing to the need for worldwide application, extreme environmental conditions and lacking long-term experience. In future, it may become a valuable method to assist various lightweight applications in different materials

The problem addressed before RAMSSES project is to demonstrate that composite (FRP, fibre reinforced polymer materials) overlamination is suitable both as a repair technology for damaged structures in a marine environment as to improve the pristine properties of welded joints.

The guidelines for developing composite patches for repair and reinforcement in shipbuilding developed in previous projects as COPATCH [1] or PARCHE [2] projects provide an acknowledged and well-proven approach that can be adopted for the initial phase of this work.

Composite patches are considered able to enhance the fatigue life of a cracked plate and there is evidence of this occurring in the aviation industry. Bonding of the composite to a metal structure has the benefit of providing an alternative load path, which by passes the stress concentration points and consequently reduces the problem of local fatigue life. This is possible due to the strength of the composite material, which increases the cross-sectional modulus of the structure, similar to the double plate effect.

The technology of overlamination of composites over metallic structures is based in the pioneering work of The Australian’s Department of Defence [3]. This new solution emerged in the seventies for repairing aircraft structures, as a quick way to make repairs minimizing unplanned downtimes.

Three types of repairs are considered for maritime components [4]: Considering this classification and the interests on which RAMSSES focuses, structural 2D repairs, which typical cases include flat patches bridging a crack, have be mainly studied for metallic materials repair cases.

The working strategy within RAMSSES project was selected and designed a structural detail of a cruise vessel, which is prone to be fatigued and then, prone to crack as the demo case.

2.       EXPERIMENTAL PROCEDURE

2.1          MATERIALS

As it has been previously mentioned, the objective of this work is to study the effect of a composite overlamination on a damaged steel structure in a vessel real case. The demonstrations were manufactured using AH36 steel plates of 8, 12 and 25 mm. The crack of the demo cases was performed by EDM in all of them.

To select the optimal composite patch for repairing the cracked steel, three composite systems are analyzed at laboratory scale manufactured by two different manufacturing techniques. Thermal and mechanical characterization were carried out for the three composite systems.

Two types of resins were studied, epoxy and vinylester, manufactured with dry carbon fibre reinforcement, as well as carbon-epoxy preimpregnated materials were be studied.

The reference of the Vinylester resin is Epovia Optimum KRF 4436 AI of Polynt Composites, of the Epoxy resin is PRIME™ 27 EPOXY INFUSION SYSTEM of Gurit, the Carbon Fibre Bidirectional Fabric is CC 200 P – 120 of KordCarbon and the Prepreg is Carbon Fibre Epoxy resin with the reference VTTM 246-42%-3KHS-2X2T-1999-1250 of SOLVAY.

2.2          MANUFACTURING PROCESS

Three composite systems and two manufacturing processes have been selected to be studied in this work. Composite characterization tests were carried out both for composite system selection and to feed the patch numerical modelling.

Two different patch manufacturing techniques were carried out. Vacuum Resin Infusion, (VRI) technique were used for dry carbon fiber and epoxy and vinylester resin. The manufacturing methodology used for prepreg was Hand-lay-up (HLU) combined with the application of vacuum after lamination (Vacuum Bag, VB) to obtain maximized mechanical properties (see Figure 1).

So, three composite systems and two manufacturing processes were selected to be studied in this work.

Figure 1. Composite laminates manufactured by VRI (left) and HLU + VB (right)

2.3          JOINING STRATEGIES

This task was carried out to evaluated at laboratory scale which surface treatment leads to a better bonded joint and analyzed the mechanical characteristics of the dissimilar joint. The aim of these tests is to select the steel to composite bond with the optimal surface conditions in order to implement at the demo scale.

Taking into account the Surface conditions that BV NI613 defines about the optimal surface preparation and the different steps of steel prior to bonding. In this work, the optimal surface conditions of steel to apply the composite were defined by single lap shear tests.

Three steel surface conditions were tested: steel without treatment as a reference (cleaned with alcohol), gritblasting and sanding.

Single lap joint specimens were manufactured using three composite systems by two different manufacturing processes direct on the steel and with the three different surface treatments mentioned above.

Figure 2. Steps of manufacturing process of the SLJ specimens by VRI (left) and HLU + VB (right)

Figure 3. SLJ coupons manufactured by VBI (left) and with the prepreg by HLU+VB (right)

2.4          REPRESENTATIVE MULTI-MATERIAL DEMONSTRATOR: SELECTION AND MANUFACTURING

After agreement by all RAMSSES partners, the final demo case of was to develop the “Structural detail on a deck opening of a cruise ship”. A draw of this type of structural detail can be observed in Figure 4.

The main objective of this demo case is to use the composite overlamination technique to prevent the crack initiation or act as crack arrestor in ship structural details. This objective is pursued in RAMSSES studying the application of a composite patch on a typical marine structural detail. Various structural details were considered, but the selected one consists on a corner of a deck opening of a cruise ship of AH36 steel (see Figure 4). In fact, these zones of the ship are typically subject to stress concentrations and consequent phenomenon of crack initiation.

Figure 4. Typical bulk carrier defect in way of structural detail, Defects in inner bottom plating due to insufficient strength in way of manhole or possible high stress concentrations (left). A draw of the Structural details of deck opening of a cruise vessel of the selected Demo case (right)

A composite patch could be a good solution for this area which presents risk of loss of structural strength, avoiding the increase in weight associated with the introduction of an extra metal sheet, as well as associated hot works (welding). According to this, composite patch design was done based on the dimensions and requirements of this structure. Likewise, for the validation of the repair / reinforcement solution, a representative medium-scale demo case was designed and tested.

Figure 5. Prospective view of the large-scale demonstrator. Final geometry and dimensions of the large-scale demonstrator

3.      RESULTS AND DISCUSSION

3.1          RESULST OF CHARACTERIZATION OF COMPOSITE MATERIALS

Characterization of the composite laminates manufactured with the different materials and different manufacturing methods is needed in order to feed the numerical calculation and modelling for patch design in order to select a system (resin / fiber / manufacturing method / substrate surface conditions) to apply the composite overlamination solution for demo case. In that sense, the tests were carried out were the following: Tensile test, as per ISO 527-4, Compression test, as per the standard ISO 14126 and Interlaminar Shear Tests according to ISO 4585 and ASTM D3518.

Specimens for tensile, compression and shear strength tests were cut in two directions (0⁰ and 90⁰) from manufactured coupons, to check if the resin infusion process was homogeneous through the entire coupons manufactured with bi-directional carbon fibers. For all the tests defined, the number of specimens defined by the correspondent standard were tested, and results were analyzed (minimum of five specimens in each case).

Figure 6. Average test results of composite characterization

Figure 6 shows the results of composite characterization. From these results, epoxy-based system has slightly higher tensile strength as well as elastic modulus than the other two systems. For compression performance, the best system would be the one manufacture with the prepreg. The main difference among these results is the one for shear performance, where epoxy-based compound is almost three times better than any of the other two composite systems.

A thermogravimetric analysis (TGA) analysis was performed to estimate the amount of resin and the fibre fraction within the manufactured composite. The results of the analysis are in Table 1.

Table 1. The results of the analysis of fibre volume fraction

Composite system	Resin content (%)	Fiber content (%)
VB-C/E CF-epoxy manufactured by vacuum resin infusion	38.9	61.1
VB-C/V CF-vinylester manufactured by vacuum resin infusion	41.9	58.0
VB-PP C/E CF-epoxy  manufactured by prepreg method	32.1	67.3

3.2          RESULTS OF MULTI-MATERIALS JOINT CHARACTERIZATION

Within the Intermediate level tests, multi-materials joint characterization was carried out. Hybrid joints were performed in order to evaluate adhesion between the composite systems and steel. Lap shear tests were done according to ASTM D5868 in order to define the bond line strength and select then the optimal material combination. Besides, the aim of the bond line characterisation tests is to provide the inputs required for the theoretical models used to predict the behaviour of the bonded repair.

An objective stiffness ratio (SR), was selected (based on previous experiences from RAMSSES partners) and the design of the joint (thickness of materials: steel and composite, overlap, etc.) was made based on it. Results from these tests gave information to choose the final composite system to be applied on the demo case.

The average values calculated are depicted in Figure 7. Taking these results into account, it can be observed that prepreg systems provide higher adhesive shear strength, as well as grit blasting surface treatments led to higher mechanical joint performance.

Figure 7. Average SLJ-test results of each combination of composite to AH36 joints

3.3          COMPOSITE PATCH DESIGN OF A REAL SCALE TEST

The objective of this test was to reproduce the real mechanical behaviour. The demo case selected was a structural detail, which is common to several types of ships (cargo ships, cruise ships, navy ships, etc.) is typically characterized by the presence of stress concentration that if not correctly mitigated can lead to phenomenon of crack initiation. For these reasons it was considerate as test case to evaluate the effectiveness of the composite patch both in preventing the crack initiation and mainly, as crack arrestor.

Several patch configurations were evaluated and finally the best patch design was selected for the building of the real scale demonstrator. The effectiveness of the patch was assessed by means of numerical simulation (FEA, finite element analysis), correlated to laboratory tests results. The numerical model was developed, and the dynamics of the fatigue test was simulated.

The patch design was performed taking into account the guidelines of reference [5] and constraints imposed by the testing machine conditions. In order to prevent debonding phenomena, the patch layers were tapered at the patch extremities. By means of the FEA analyses to identify the optimum SR of 0.28, the optimum patch thickness, tp, of 13.74 mm and the optimum fibre orientation in each group of layers (at 0°and 45°). In addition, the effectiveness of the patch selected as crack arrestor, the fracture parameters at the crack tip were evaluated. Finally, an optimal length crack of 40mm was selected by FEA analyses and was performed in the demo.

Figure 8. Composite patch manufactured by HLU and VB techniques for demonstrators. Optical fibers were embedded during the manufacturing process in order to monitor the real fatigue test of the demonstrators.

3.4          FATIGUE TEST OF REAL DEMO CASE

Demonstrators testing were tested in a universal machine, under similar loads as the original structure experiment in service. Therefore, the final demo case (patched one) ready to test is represented in the Figure 9.

Figure 9. Patched Demonstrator and strain gages sensors positions

Figure 10 shows the final set up carried out to test the demonstrators, the three Patched and the other three UnPatched demo cases. Finally, the fatigue tests were developed in a RUMUL 250KN machine.

Figure 10. Final set up for fatigue testing of demo cases; UnPatched (left) and Patched (right)

The fatigue test was monitored with strain gages and fiber optical sensors. Figure 11 shows the response of the demo case UnPatched under fatigue conditions. It can be seen the relationship of the strain at different distances of the crack tip with the number of cycles applied and the crack growth.

Therefore, it can be observed that the strain gage placed at 5mm of the crack tip (position 1) measure overload when the crack cross the gage, so this is a signal that the crack length was 5mm at this time.

Moreover, the strain gage placed at 30 mm of the crack tip (position 2) measure maximum strain when the crack cross the gage, so this is a signal that the crack length was 15 mm after 15000 cycles.

Regarding the FBGs response, the FBGs located at 5mm from the end of the crack was out of range during the preload process, before to start with the cycles. This means that the crack started to growth during this preload process. The FBGs located at 25mm from the end of the crack, growths continuously, in line with the crack growth, while the FBGs located at 55mm, detect the crack growths around the 15000 cycles, as the gauge at the same position.

Figure 11. Behaviour of the Cracked UnPatched demo case measured with Strain gages (left) and FBGs (right) on fatigue testing

Figure 12 shows the response of the demo case Patched under fatigue conditions. It can be seen the relationship of the strain at different distances of the crack tip with the number of cycles applied and the crack growth.

Therefore, it can be observed that the strain gage placed at 5mm of the crack tip (position 1) measure overload when the crack cross the gage, so this is a signal that the crack length was 5mm at this time. Moreover, the strain gage placed at 30 mm of the crack tip (position 2) measure overload when the crack cross the gage, so this is a signal that the crack length was 30 mm after 100000 cycles.

Regarding the FBGs response, in this demo one extra FBGs was located over the end of the steel crack (in the steel side without patch, as the gauges side). These FBGs (black signal) detects that the crack started to growth with the first load cycle. The crack pass through these FBGs at 800 cycles. The signal of the FBGs located at 5mm, embedded between the steel and the first CF layer or between CF plies (in the middle CF ply), growth continuously up to the 60000 cycles and then still growth but slowly. While the FBGs located at 25mm from the crack, after the 60000 cycles increase more the signal, in accordance with the gauges signal, because the crack reach the 25mm position. Finally, the FBGs located at 55mm from the crack shows a jump in the signal around the 110000 cycles according to the associated gauge signal. This means that the crack have overlapped the 25mm position and is arriving to the 55mm position.

Figure 12. Behaviour of the Cracked Patched demo case measured with Strain gages (left) and FBGs (right) on fatigue testing

Finally, Figure 13 shows pictures about the crack growth on the bottom side of the P5 plate.

   Figure 13. Crack growth in fatigue test of demo case at the same number of cycles, demo Patched (left) and UnPatched (right).

As conclusion, at the end of the fatigue test on the cracked UnPatched demo case, the crack has been increased its length on 30 mm after 37500 cycles. On the other hand, regarding the cracked patched demo case, the crack has been increased its length on 30 mm after 100000 cycles. Therefore, this behaviour represent an improvement of the 2.6 times of the fatigue life.

4.       CONCLUSIONS

The demo case was selected as a representative of a structural detail in a cruise vessel made on AH36 steel. The final demo design and the best patch configuration were selected by FEA calculations. Six demonstrators were built. The crack was performed by EDM in all of them. Three demonstrations of them have been Patched and three of them have been UnPatched. Composite manufacturing, curing and post-curing, were monitored by FBGs embebed among the layers during the manufacturing process. Demo testing were performed by fatigue test and monitoring by strain gauges and FBGs positioning and distribution in the high stress areas along the demonstrator. Fatigue tests were carried out in the six demos to compare crack evolution in cracked patched and un-patched demonstrators.

As conclusion after the fatigue tests, composite patch works as crack arrested. Therefore, this behaviour represent an improvement of the 2.6 times of the fatigue life comparing UnPatched and Patched damaged steel.

Bibliography

[1] COPATCH project. Project ID: 233969; SCP8-GA-2009-233969; FP7-SST-2008-RTD-1

[2] PARCHE Project results. E6.2. Caracterización mecánica y monitorización de demostradores (EXP 00064277 / ITC‐20133012), CARDAMA. SPAIN. February 2015

[3] A, Baker, et al. Advances in the Bonded Composite Repair of Metallic Aircraft Structure. Elsevier. 2012

[4] RECOMMENDED PRACTICE DNV – RP – C301, Design, Fabrication, Operation and Qualification of Bonded Repair of Steel Structures, April 2012

[5] “Determination of Structural Capacity by Non-linear FE analysis Methods”, recommended practice DNV. DNV-RP-C208. June 2013.

Acknowledgments

RAMSSES project has received funding from the European Union’s Horizon 2020 research and innovation programme within the framework of Mobility for Growth, Innovations for energy efficiency and emission control in waterborne transport, under grant agreement No 723246. Thanks to CETENA, GALVENTUS, and CARDAMA Shipyard to participate in this task, as well, for their support and hard work.