Displacement rate effect in the fracture toughness of glass fiber reinforced polyurethane

Abstract

Composite structures currently used in the oil industry must meet strict requirements for design and safety reasons. They need to maintain strength under varied displacement rates throughout its lifetime. It is therefore critical to fully understand the fracture behavior of such composites. This work presents experimental results regarding the influence of a range of displacement rates on the fracture energy in mode I, G_Ic, of glass fiber reinforced polyurethane used in the oil industry to repair and reinforce pipelines with corrosion damage. To determine G_I_c as a function of displacement rate, double cantilever beam specimens were tested, with displacement rates of 2, 20 and 200 mm/min with different thicknesses. A complementary numerical study was performed with the aim of predicting strength using the measured values. This work has demonstrated a significant influence of the strain rate and composite thickness on G_IC of the composite materials, with higher rates and thicker specimens causing an increase in the G_IC values.

Keywords

Polymer-matrix composites fracture toughness mechanical testing

Introduction

One of the most dangerous problems faced by the oil and gas industry is the corrosion of piping and pipelines, which can greatly compromise its integrity, leading to leakages or explosions due to high pressure of the fluid being transported.^1,2 The application of polymeric composites as a material for repairing such critical infrastructures has continuously increased in the last decades.³ An example is the fiber reinforced polymer (FRP) matrix composites, which provides some advantages, namely: short time required to complete a repair; allows product transmission in the piping system without compromising its efficiency, generally enables an extension in lifetime of up to 20 years, and offers a safe procedure since no heat or sparks occur during its application.⁴

A composite material increasingly used to repair corroded piping and pipelines is the resin-impregnated fabric bondages, being the most common the knitted fiberglass fabric which is impregnated with a polyurethane resin.^5–7 Among this type of materials, the glass fiber reinforced polymer (GFRP) matrix composites is one of the most capable, due to its load-bearing capabilities in pipelines reinforcement, and this type of rehabilitation is becoming well accepted as an engineering practice since it also reduces the corrosion process and shields from the external environmental conditions.^8–11 The application of GFRP offers: short curing time with the presence of water, high abrasion resistance, significant tear strength, excellent shock absorption, flexibility in its application/casting around the pipes an elasticity.^12–14

Since the application of GFRP as a repairing and reinforcing material of pipelines is mainly structural, it is therefore important to understand and be able to predict its mechanical behaviour under several circumstances, in order to guarantee an accurate and appropriate intervention. The variation of the mechanical properties of GFRP as a function of strain rate is an issue which must be addressed,^15–17 as the increase of the strain rate leads to higher moduli, since the polymer chains have decreased the relaxation time.¹⁸ For short periods of time, the polyurethane molecules, not having the capability to reorient, will probably react to a stress by distorting intramolecular distances, therefore, resulting in high energy effect leading to a high modulus.¹⁹

An important mechanical property that plays an important role in the integrity of the repaired pipes and pipelines is the fracture toughness in mode I, G_IC, of the GFRP introduced, especially with the variation of the strain rate. The vast majority of tests performed with the aim to characterize the material G_IC are performed through the use of double cantilever beam (DCB) specimens.^20–26 The two main standards for this test are the ASTM D3433-99 and ISO 25217.²⁴

A numerical study of mode I interlaminar crack growth using fiber reinforced composite DCB specimens was conducted by Todo et al.,²⁷ being stated that crack-tip bluntness induces a stress concentration along the crack form, causing the stress points to move from the tip to the corner along the crack form, and originating crack growth towards the fiber/matrix interface. It was also stated that the presence of high tensile stresses in the region of the interface near the insert film (initial crack length) is, therefore, an indication for this type of specimens, and that if the fiber/matrix interface is not strong enough, a crack may start at the interface and not in the region of the inserted film.

The value of G_IC of glass fiber reinforced polyester composite, under quasi-static conditions, was performed by Leonard et al.,²⁸ by using compact tension (CT) specimens. The influence of fiber volume fraction in the composite matrix was considered, revealing a considerable increase in the G_IC values and critical energy release rate with the increase of the fiber content. A 2D numerical simulation was also performed by the authors with a good accuracy.

Testing DCB specimens of glass fiber reinforced vinyl ester under quasi-static conditions, Tucker et al.²⁹ found that an increase of G_IC of the composite occurred after a short period of post-cure (the same behaviour was observed for the resin G_IC), and that after a prolonged post-cure such value would continuously decrease (opposite to what found for the resin G_IC behaviour), which was believed to be caused by the weakening of the fiber/matrix bonding strength.

Although not a lot of work was performed regarding the increase of strain rate, reference values of the influence of an impact load in G_IC of GFRP materials can be found in the literature. Chen et al.³⁰ determined the mechanical properties of a polyurethane pultruded composite, under quasi-static and impact conditions. The authors found that the fiber content plays an important role and that the assessed mechanical properties increased with the increase of the fiber content for the case of quasi-static conditions; as for the case of impact loads, the dynamic mechanical moduli revealed to be higher than those of the matrices and the influence of the fiber content was the same as observed for quasi-static conditions.

Jar et al.³¹ tested the GFRP specimens under impact, observing that the energy absorbed by the specimens during the test was independent of the variation of the matrix toughness. Also, the impact toughness is believed to vary with the composite matrix toughness if the impact toughness is defined as the energy absorbed per unit of area damaged. Kuboki et al.³² performed a comparison between the delamination resistance under several impact speeds between two GFRP materials (polyurethane matrix and isophthalic matrix) and using DCB specimens, revealing that the composite with polyurethane matrix presented less influence of damage size and higher impact toughness; however, in terms of energy absorbed between both materials and for the testing impact speeds, the values were similar for all the testing conditions.

An experimental procedure followed by a numerical analysis of the failure criteria for mixed mode delamination of GFRP materials was performed by Marat-Mendes et al.,³³ namely, in mode I, mode II and mode III. The specimens used to experimentally determine the value of G_Ic were the DCB, and by using the virtual crack closure technique (VCCT) it was possible to characterize G_Ic of the composite, revealing that by using an appropriate failure criterion³⁴ it is possible to achieve a close prediction of the materials behaviour.

This work assesses the influence of a range of displacement rates on the fracture energy in mode I, G_Ic, of glass fiber reinforced polyurethane specifically used in the oil industry to repair and reinforce pipelines with corrosion damage. Since oil moves through pipelines at different pressure conditions,³⁵ it is important to understand the variables which contributes to the design of composite repair and reinforce systems. Usually standards consider a fixes flow pressure but this does not occur in the field. To determine G_Ic as function of displacement rate, double cantilever beam specimens were tested, with displacement rates of 2, 20 and 200 mm/min with different thicknesses.

Experimental setup

Materials

The composite material used in this research was the Syntho-Glass XT®, from NRI (Neptune Research Inc.), which is a polyurethane pre-impregnated, bi-directional E-glass fiber composite used to repair and reinforce internal and external corrosion on pipeline or structures. This product is water-activated polyurethane resin, which reduces composite preparation time by 50%. Previous studies^1,2,36 showed the behavior of this material when tested at different strain rates and temperatures. The properties of the composite adherend were provided by the manufacturer and are presented in Table 1.

Table 1.

Composite properties.¹

Property	Syntho-glass XT
Circumferential tensile modulus	30.5 GPa
Axial tensile modulus	15.8 GPa
Flexural strength (MPa)	405
Tensile strength (MPa)	234
In-plane shear strength (MPa)	51.9
Hardness (Shore D @ 24 C)	83 (24 h)

The composite laminate of dimension 100 × 300 mm² was fabricated from stacking by hand lay-up technique. A non-adhesive insert (Teflon sheet) of thickness less than 75 µm was introduced as a pre-crack in the mid-plane of the laminate. The thickness of each Syntho-Glass XT® sheet is 0.33 mm, then three different laminate thickness were manufactured, 4 plies (1.32 mm), 10 plies (3.3 mm) and 20 plies (6.6 mm) thick. All the laminates were cured at room temperature for 24 h. After removal from the press, all the samples were cut from the laminates using a diamond saw.

The GFRP specimen geometry was based on ASTM-D 5528-13,³⁷ which is used to determine the interlaminar fracture toughness of fiber reinforced plastics in mode I. Specimens have 140 mm length, 25 mm width and 1.32, 3.3 and 6.6 mm thickness. The previously introduced Teflon sheet produced a pre-crack of a₀ = 35 mm to induce stable crack propagation.

In order to avoid any bending failure from the composite, 12.5 mm thick steel plates were bonded to the composite surfaces to achieve a DCB test configuration as shown in Figure 1.

To improve the adhesion between the steel plate and the composite, both materials were surface prepared. The steel plates were blasted with 180 mesh sand grit and were cleaned by wiping with an acetone-soaked cloth.

The composite samples were treated with an Arcotec (Rotweg, Germany) Arcojet PG051 oxygen plasma device. Plasma surface treatment has been known very effective because the plasma surface treatment increases the surface free energy of composite adherend, which enhances the wetting between the adherend and the adhesive.³⁸ The laminates were placed on a wood frame, isolating the sample and allowing the plasma to complete interact with the composite. Exposure time of 3 min was chosen. The contact angles of the water drop on both the plasma surface non-treated composite and treated one are shown in Figure 2.

The contact angle of the liquid drop on the composite decreased after the plasma surface treatment because the surface free energy of the composite site increased.

Method

The DCB specimens were tested according to ASTM D5228³⁷ at different crosshead displacement rates, 2, 20 and 200 mm/min, respectively.

Three DCB specimens were manufactured for each composite thickness and the tests were performed in an INSTRON model 3367 universal test machine (Norwood, Massachusetts, USA) with a load capacity of 30 kN at room temperature. Six specimens were tested for each displacement rate. Figure 3 displays the experimental setup for DCB testing.

Results and discussion

The determination G_Ic for the DCB specimens was accomplished using the compliance-based beam method (CBBM)²¹ which only uses the load and displacement curves of a DCB test to determine the G_Ic. Contrarily to other methods such as compliance calibration method and (CCM) and the corrected beam theory (CBT),^20,21 the measurement of the crack length is not necessary. Measuring the crack length during the test procedure affects the accuracy of the results since the crack evolution is instable for brittle materials. The CBBM is a relatively straightforward but robust method, based on an equivalent crack length (a_eq), and it only depends on the specimen’s compliance during the test. Applying this to the DCB test specimen gives

G_{Ic} = \frac{6 P^{2}}{B^{2} h} (\frac{2 a_{eq}^{2}}{h^{2} E_{f}} + \frac{1}{5 G})

(1)

where P is the applied load, B is the specimens width, h is the thickness and a_eq is an equivalent crack length estimated from the current specimen compliance and taking into consideration the damage zone, E_f is a corrected flexural modulus to account for stress concentrations at the crack tip and stiffness variability between specimens. G is the shear modulus of the adherends.

Figure 4 shows a representative experimental load–displacement (P–δ) curve of the DCB specimens as a function of composite thickness for 2 mm/min.

As it can be seen at the beginning of the P–δ curves, the evolution of the curve is linear, and the behavior is assumed to be reversible. A loss of linearity is observed corresponding to the appearance of an irreversible process in the composite, such as plasticity or damage. After the peak load is reached, the applied force decreases progressively, which is related to a stage of crack propagation. At the end of the test, a rupture is finally observed when the crack front is close to the free edge of the specimen.

According to Figure 4, stiffness remains unaltered as the composite thickness increase. The initial P–δ curves present similar slopes. Also, by increasing the composite thickness, the strength increases to a maximum, which is observed with a thickness of 6.6 mm. This behavior is observed for all composite thicknesses investigated. In addition, analyzing the curves and monitoring of cracking during testing, slight fluctuations in propagation velocity can be observed, which are indicative of possible heterogeneities within the composite layers and a brittle behavior.

Data points from fracture toughness experiments were used to construct a delamination resistance curve (R-curve), which is the graph of fracture toughness (G_Ic) versus equivalent crack length (a) for DCB specimens with composite thickness of 1.32, 3.3 and 6.6 mm is presented, respectively in Figure 5. R-curves are used to identify the fracture energy from the plateau corresponding to crack propagation.

The values considered were taken from the stabilization value, measured in the stabilized section of the R-curve of each specimen tested. As a result, the increase of the composite thickness and crosshead displacement speed leads to an increase of the fracture energy.

Table 2 displays the average values of the fracture energy, G_Ic, as function of the different composite thickness and crosshead displacement rates.

Table 2.

Fracture toughness in N/mm (avg ± St. dev.) for different composite thickness and crosshead displacement rates.

Displacement rate (mm/min)	Composite thickness (mm)
Displacement rate (mm/min)	1.32	3.3	6.6
2	0.63 ± 0.08	1.19 ± 0.10	1.81 ± 0.09
20	0.75 ± 0.05	1.31 ± 0.10	2.10 ± 0.14
200	1.35 ± 0.12	2.08 ± 0.26	2.70 ± 0.12

From Table 2, it can be reported that larger composite thickness produces higher fracture toughness. Increasing the composite thickness when DCB specimens were tested at 2 mm/min displays an elevation of 88.9% (3.3 mm) and 187.3% (6.6 mm) when compared to specimens with 1.32 mm thick. Elevating the crosshead speed to 20 mm/min, the fracture toughness of the 3.3 mm and 6.6 mm thick composites presents an increase of 74.6% and 180.0%, respectively, when the fracture toughness of composites with 1.32 mm thick is calculated. The same behavior is reported for specimens tested at 200 mm/min. Higher thickness displays higher fracture toughness. An increment of 56.4% and 103.0% is observed when composite thickness increases from 1.32 mm to 3.3 mm and 6.6 mm, respectively. These results showed that the fracture toughness depends on the composite thickness.

When comparing the influence of crosshead displacement for the same composite thickness, similar behavior is observed. Increase in the crosshead displacement rate increases the glass fiber reinforced polyurethane fracture toughness. By increasing the crosshead displacement, speed higher fracture toughness is calculated. Analyzing the 1.32 mm thick composite, an increase of 19.0% is calculated when the speed is elevated from 2 to 20 mm/min and 111.1% higher when 200 mm/min is compared; 10.1% and 74.8% higher fracture toughness is calculated for the 3.3 mm thick composite when DCB specimens were tested at 20 mm/min and 200 mm/min, respectively, compared to 2 mm/min. Finally, an increase of 16.0% for specimens tested at 20 mm/min and 49.2% for 200 mm/min when compared to 2 mm/min. The effect of composite thickness and crosshead speed on the fracture energy, G_IC, can also be observed in Figure 6.

According to Figure 6, by increasing the crosshead displacement speed, a linear evolution of the fracture toughness is observed for all composite thicknesses tested. Also, as the composite becomes thicker, the fracture toughness increases as well. The maximum value of G_Ic is obtained for a composite plate with 20 layers, 6.6 mm thick. The linear behavior presented by the fracture toughness can be represented by equation (2)

G_{Ic} (s) = as + b (s)

(2)

where a is a constant and b(s) is the parameter that depends on the crosshead speed (s). According to the parameter presented in Figure 6, a is a = 0.004 and b can be represented by a power law b(s) = 0.57 s^0.23. Table 3 presents the comparison between the proposed equation for the fracture toughness and the experimental test results.

Table 3.

Comparison between proposed model and experimental test result.

Composite thickness (mm)	Displacement rate (mm/min)	Fracture toughness (N/mm)
Composite thickness (mm)	Displacement rate (mm/min)	Experimental results	Proposed equation	Variation (%)
1.32	2	0.63	0.68	+7.9
	20	0.75	0.75	0
	200	1.35	1.47	+8.9
3.3	2	1.19	1.14	−4.2
	20	1.31	1.22	−6.9
	200	2.08	1.94	−6.7
6.6	2	1.81	1.94	+7.2
	20	2.10	2.01	−4.3
	200	2.70	2.73	+1.1

According to Table 3, it can be evaluated that the experimental results are within or close to the analytically predicted ranges for the studied crosshead speed. The maximum variation observed was 8.9% for composite with 1.32 mm thickness at 200 mm/min.

Analyzing the crack propagation, it can be seen in Figure 7 that the fiber bridging phenomenon occurs in the glass fiber reinforced polyurethane composites. In order to identify the fiber bridging mechanism in delamination interface, the crack view of the interface during delamination propagation is presented.

Fiber peeling from the resin as an evidence of fiber bridging and matrix cracking during delamination propagation are the mechanisms of energy absorbing during the crack propagation. Figure 8 presents the typical failure observed in glass fiber reinforced polyurethane DCB tested specimens.

As illustrated in Figure 8, the fiber bridging zone is approximately evenly distributed through the entire specimens. Experimental investigations of DCB laminates reveal that the crack jumping occurs, and the crack tends to migrate during the delamination propagation. Also, the results showed that this phenomenon can affect the resistance curve due to wavy crack path in cross-ply laminates.

Conclusions

The main goal of this study was to investigate the influence of the crosshead displacement rate on the fracture behavior of DCB specimens of glass fiber reinforced polyurethane used in the oil industry as a repair system. In order to eliminate the effect of bending moment and therefore not measuring mode I fracture toughness, thick bars were attached to the composite. Afterwards, the fracture tests based on ASTM standard were conducted on DCB specimens and the interlaminar fracture toughness was characterized based on the CBBM method. It was found that, for the range of crosshead displacement rates investigated, the fracture toughness varied linearly. The higher the crosshead displacement rate translated into, higher the value of fracture energy. That increment was also observed when thick composite laminates were studied. Thicker composites deliver higher resistance to crack propagation. It was observed that more energy was needed for fibers peeling of thicker specimens and this phenomenon may affect the initiation and propagation values of the interlaminar fracture toughness. In addition to bridging phenomenon, the matrix cracking during the delamination propagation affects the propagation toughness values of the DCB specimens. The investigation contained herein constitutes an important step towards the establishment design and repair of pipelines with corrosion damage used in the oil industry.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The authors thank the Research Foundation of the State of Rio de Janeiro (FAPERJ), the Brazilian National Council for Scientific and Technological Development (CNPq) and Coordination for the Improvement of Higher Education Personnel (CAPES) for supporting part of the work presented here.

ORCID iDs

JML Reis

JJM Machado

RJC Carbas

References

Andrade

, et al. A temperature-dependent global failure criterion for a composite/metal joint. Compos Part B 2018; 137: 278–286.

Junior

Reis

da Costa Mattos

. Polymer-based composite repair system for severely corroded circumferential welds in steel pipes. Eng Fail Anal 2017; 81: 135–144.

Schoene

. Testing of composite repairs according to ISO and ASME standards and beyond. Pipeline Sci Technol 2018; 2: 123–134.

Lim

, et al. An overview of corroded pipe repair techniques using composite materials. Int J Mater Metall Eng 2016; 10: 19–25.

Wirpsza

. Polyurethanes: chemistry, technology, and applications, UK: Ellis Horwood, 1993.

Dodge

. Polyurethane chemistry, Pittsburgh: Bayer Corp., 1999.

Bayer

. Polyurethane application research department, Leverkusen: Bayer-Polyurethanes, 1979.

da Costa Mattos

, et al. Analysis of a glass fibre reinforced polyurethane composite repair system for corroded pipelines at elevated temperatures. Compos Struct 2014; 114: 117–123.

Mableson

, et al. Refurbishment of steel tubular pipes using composite materials. Plast Rubber Compos 2000; 29: 558–565.

10.

Sathishkumar

Satheeshkumar

Naveen

. Glass fiber-reinforced polymer composites – a review. J Reinf Plast Compos 2014; 33: 1258–1275.

11.

Da Costa

RRC

de Medeiros

Ribeiro

, et al. Experimental and numerical analysis of single lap bonded joints: epoxy and castor oil PU-glass fibre composites. J Adhes 2017; 93: 77–94.

12.

Chiou

Schoen

. Effects of crosslinking on thermal and mechanical properties of polyurethanes. J Appl Polym Sci 2002; 83: 212–223.

13.

John

Bhattacharya

Turner

. Characterization of polyurethane foams from soybean oil. J Appl Polym Sci 2002; 86: 3097–3107.

14.

Desai

, et al. Effect of polyols and diisocyanates on thermo-mechanical and morphological properties of polyurethanes. Eur Polym J 2000; 36: 711–725.

15.

Jacob

, et al. Strain rate effects on the mechanical properties of polymer composite materials. J Appl Polym Sci 2004; 94: 296–301.

16.

Jacob

, et al. The effect of loading rate on the fracture toughness of fiber reinforced polymer composites. J Appl Polym Sci 2005; 96: 899–904.

17.

da Costa-Mattos

, et al. An alternative methodology to repair localized corrosion damage in metallic pipelines with epoxy resins. Mater Des 2009; 30: 3581–3591.

18.

Shaw MT and MacKnight WJ. Introduction to polymer viscoelasticity. Hoboken, NJ: John Wiley & Sons, 208.

19.

Saunders JH and Frisch KC. Polyurethanes: chemistry and technology. New York, NY: Wiley, 1962.

20.

Campilho

. Repair of composite and wood structures, Portugal: Instituto Superior de Engenharia do Porto, 2009.

21.

de Moura

Campilho

Gonçalves

. Crack equivalent concept applied to the fracture characterization of bonded joints under pure mode I loading. Compos Sci Technol 2008; 68: 2224–2230.

22.

da Silva

. Handbook of adhesion technology, Berlin: Springer Science & Business Media, 2011.

23.

Banea

da Silva

. Mechanical characterization of flexible adhesives. J Adhes 2009; 85: 261–285.

24.

Carlos MS Canto. Strength and Fracture Energy of Adhesives for the Automotive Industry, Master Thesis, Faculdade De Engenharia Da Universidade do Porto Oporto, Portugal, 2013.

25.

Brunner

Blackman

Davies

. A status report on delamination resistance testing of polymer–matrix composites. Eng Fract Mech 2008; 75: 2779–2794.

26.

Machado

, et al. Mode I fracture toughness of CFRP as a function of temperature and strain rate. J Compos Mater 2017; 51: 3315–3326.

27.

Todo

Jar

P-Y

. Study of mode-I interlaminar crack growth in DCB specimens of fibre-reinforced composites. Compos Sci Technol 1998; 58: 105–118.

28.

Leonard

, et al. Fracture behaviour of glass fibre-reinforced polyester composite. Proc Inst Mech Eng Part L 2009; 223: 83–89.

29.

Tucker

Compston

Jar

P-Y

. The effect of post-cure duration on the mode I interlaminar fracture toughness of glass-fibre reinforced vinylester. Compos Part A 2001; 32: 129–134.

30.

Chen

C-H

C-CM

. Pultruded fibre-reinforced polyurethane composites. III. Static mechanical, thermal, and dynamic mechanical properties. Compos Sci Technol 1994; 52: 427–432.

31.

Ben Jar

P-Y

, et al. Evaluation of delamination resistance of glass fibre reinforced polymers under impact loading. J Adv Mater 2000; 32: 35–45.

32.

Kuboki

Jar

P-Y

Forest

. Influence of interlaminar fracture toughness on impact resistance of glass fibre reinforced polymers. Compos Sci Technol 2003; 63: 943–953.

33.

Marat-Mendes

Freitas

. Failure criteria for mixed mode delamination in glass fibre epoxy composites. Compos Struct 2010; 92: 2292–2298.

34.

Sørensen

Jacobsen

. Characterizing delamination of fibre composites by mixed mode cohesive laws. Compos Sci Technol 2009; 69: 445–456.

35.

Ganat

Hrairi

Hawladera

MNA

, et al. Development of a novel method to estimate fluid flow rate in oil wells using electrical submersible pump. J Petrol Sci Eng 2015; 135: 466–475.

36.

Reis

JMLD

. Displacement rate effect on the flexural properties of glass fiber reinforced polyurethane. Mater Res 2016; 19: 183–186.

37.

ASTM D5528-13. Standard test method for mode I interlaminar fracture toughness of unidirectional fiber-reinforced polymer matrix composites. West Conshohocken, PA: ASTM International, 2013.

38.

Johansson

. Surface modification of plastics. Appl Plast Eng Handbook 2017; 1: 443–487.