Effect of interactions of two holes on tensile behavior of patch repaired carbon/epoxy woven laminates

Gursahib Singh Bhatia, Arockiarajan A.

Department of Applied Mechanics, Indian Institute of Technology Madras, 600036, Chennai, India

Keywords:Patch repair Tensile Digital image correlation Woven Multiple damages

ABSTRACT The conventional case of patch repair involves bonding a patch over single damage/hole in the laminate.This work investigates the effect of interaction of two holes on the tensile behavior patch repaired carbon epoxy woven laminates. The specimens of [0°/45°/45°/0°] laminates were repaired with adhesively bonded two-ply [45°]2 external patches. Three different cases of drilled specimens were produced with different hole arrangements viz. specimens with single central hole (SH), with two holes aligned along the longitudinal axis (LH) and with two holes along transverse axis (TH). The two-hole specimens were repaired with two different types,i.e. single large patches(SP)and with the two smaller patches(DP) of combined bonding area equal to the single large patches. Digital image correlation (DIC) was employed to capture strain contours. The results reveal the difference in the load transfer through the patches depending upon the arrangement of holes. The TH repaired specimen exhibit significant load recovery(SP-32.75%, DP-34.62%) while the LH specimens result in very marginal (SP- 6.11%, DP-4.10%) recovery compared to their drilled case.The TH specimen failed by crack growing through both the holes beneath the patch, while the LH specimens failed by the failure through only one hole. The use of single large patch over multiple holes and multiple small patches individually over each hole has no significant influence on load recovery.

1. Introduction

Composite laminates offer several advantages over metal alloys such as high strength/stiffness to weight ratio, better corrosion resistance, easier tailoring of mechanical properties etc. The advantages offered by composites make them a befitting replacement of the alloys in numerous applications ranging from automotive,aircraft, naval structures, wind turbines, sports equipment, prosthetic, armours etc. [1-5]. However, the composite laminates are inherently prone to the damages arising due to impact events.During the service life a structure may experience impact due to various incidents such as bird strike or runway-debris strikes,hailstorms, tool drops during maintenance of the structure or ballistic events in combat environment.These impact incidents induce damages in the form of fiber breakage, matrix cracking, delamination etc. [6-8]. This results in degraded structural performance of the laminate and affect the overall integrity of the structure[9-11]. The performance of the structure is restored either by replacement of the damaged laminate or by performing a suitable repair operation over the damaged zone. The replacement of the panel may not always be an economically viable solution due to high costs involved in replacement. Hence, repair in the form of bonded patch repair presents a suitable alternative to restore the performance [12,13].

The bonded repair can be further classified into two primary categories viz.scarf repair and external patch repair.The process of scarf repair involves removal of material from the location of damage in the form of a tapered hole and then bonding a patch of matching taper over it.This is a complex process and requires high skilled labor and sophisticated equipment to perform the repair.On the other hand, the external patch repair involves removal of damaged material in the form of cylindrical hole and then bonding the patches over the hole. Though the external patch repair is a relatively simpler technique from manufacturing aspect, the postrepair performance is however governed by various parameters which need to be designed [14]. These parameters range from stiffness,size,thickness and shape of the patch,adhesive properties etc.

Cheng et al. [15,16] investigated the effect of stiffness of the patches by employing patches with plies of different orientations.They reported a change in damage initiation zones based upon the stiffness of the patch. The material for the patch is also an important parameter which governs the stiffness as well as the adhesion of the patch. Jefferson et al. [17] reported that a patch with equal fraction of glass and Kevlar fibers present the most favorable patch type in post repair tensile response. The presence of two different types of fibers in patch utilize the high tensile stiffness of Kevlar fibers and better adhesion property of glass fiber. In another work[18]they compared the inter ply hybrid patches with the intra ply hybrid patches and reported that the intra ply hybrid patches exhibit the better post repair performance. Another important parameter is the shape of the patch.It influences the distribution of the stresses around the repaired zone. Kashfuddoja et al. [19] reported the effect of in-plane patch shapes on the stress concentration factor (SCF) and peel stress around the repair zone.Moreover, in case of the external patches, the variation of ply dimensions in the patch laminate results in patches with different shapes at their ends across the thickness.Lee et al.[20]reported the influence on the bonding strength due to these different patch end geometries. The patch size influences the development of shear stresses and the plastic zone in the adhesive in the repaired zone[21,22]. Soutis and Hu [22] reported that the plastic zone gets influenced by the increase in size of the patch up to an optimum patch size, beyond which the patch size does not influence the plastic zone.The placement of the patch on either single side of the parent laminate or both the sides of the parent laminate also has an influence on post repair performance. Kashfuddoja et al. [23] reported a higher gain in strength for double sided repair compared to single sided repair under tensile loading. The authors in their previous work [24] investigated under the flexural loading, the effect of placement of single sided patch on tensile and compressive side of the laminate and compared it with double sided patch.Single sided patch bonded on the compressive side of the laminate under flexural loading showed a very high strength recovery.Furthermore, the adhesive plays a critical role in the load transfer between the parent laminate and the patch.A very thin layer of the adhesive results in weak and brittle bonding, whereas a large thickness of adhesive results in plastic deformations and influences the load transfer.Thus there lies an optimum thickness of adhesive layer for effective load transfer [25].

From the literature it is perceived that numerous studies have been conducted to understand the influence of these parameters.The majority of the investigations on repaired composites have been conducted for a test case where the damage is assumed to be at single location in the laminate. In actual service, the laminate may experience multiple damages occurring simultaneously at different locations [11]. Few of the published works have also reported the difference in mechanical behavior of drilled laminates with multiple holes[26,27].However,the authors are not aware of the works investigating the effect of interaction between multiple holes on the post repair performance of the laminates. The repair with multiple holes presents its own unique considerations such as a selection between single large patch for multiple damages which occur at close proximity or small patches bonded individually over each damage/hole.

Thus in this work an attempt has been made to understand the interaction between two holes repaired with single and two patches(dual patch)along with the influence of the arrangement of holes in two different manner with respect to tensile loading direction. The case of two holes repaired with the two patches presents the inceptive case of employing the multiple patches for a multiple damaged system. However, for more than two holes, the multiple patches could be equal or less than the actual number of the holes.The understanding obtained from this work will provide a direction towards a more detailed further analysis of the interaction between multiple patches and the holes.

2. Experimental procedure

2.1. Materials and laminate preparation

The laminates were prepared using 400 GSM plain woven BhorForce? PC402H carbon fabric and Araldite LY5556-Ardur HY951 resin-hardener as matrix using hand layup. The properties of the fabric and resin-hardener used are presented in Table 1 and Table 2 respectively. The layups of the laminates prepared and processing conditions for the parent specimens and the patches are specified in Table 3 and a schematic representation of the layups is illustrated in Fig. 1(b). The parent laminate consists of four plies with fibers oriented at 0°in two exterior plies, while at 45°in the two interior plies.This quasi-isotropic layup allows the repair of the primary load bearing ply(0°)of the parent laminate when repaired with bonded external patches. The patches of [45°]2stacking sequence were employed for repair, as [45°]2patches have shown to recover the highest tensile strength in a previous work by authors [28]. The specimens and patches of the required dimensions were cut from the laminates by employing abrasive waterjet cutting technique (see Fig.1).

Table 1 Fabric and fiber specifications.

Table 2 Properties of resin-hardener used.

Table 3 Laminate fabrication and processing details for both patch and parent laminates.

2.2. Specimen and repair configurations

Fig. 1(a) illustrates schematic representation of all the tested specimens.As a common practice for the external patch repair,the impacted region from the damaged site is removed in the form of a hole. During the repair, the hole is filled with a filler material and the patches of required configurations are bonded over it [29]. In this work, a similar procedure was employed, wherein the hard cured two-ply patches ([45°]2) were bonded over the holes symmetrically on the both side of the parent laminate by using the same Araldite LY5556-Ardur HY951 mix as the adhesive.The drilled specimens consist of three different configurations viz. single hole at the center (SH), two holes along the longitudinal axis (LH) and two holes along the transverse axis (TH). The multiple hole arrangements (LH and TH) represent the cases of damages at more than one site. The repair consisted of single patch (SP) and dualpatches (DP) for each of the LH and TH specimens. The contact area for SP and DP repaired configurations with the parent specimens is kept equal as depicted in Fig.1. For example, in LH_SP, a single patch of size 60 mm×15 mm is bonded over the two holes.Whereas for LH_DP specimens,two patches of size 30 mm×15 mm are bonded individually over each hole along the longitudinal axis,one next to the other. In a similar manner, the equal contact area arrangement of single patch(TH_SP)and multiple patches(TH_DP)is done for the specimens with holes along the transverse axis.

2.3. Test procedure

The tests were conducted in accordance with ASTM D5766/D5766M-11[30].An electromechanically-driven UTM with 100 kN load cell capacity (see Fig. 2) was employed under displacement control with cross-head rate of 2 mm/min. DIC was employed to capture the strain. The preparation of the specimen for the DIC involves making a speckle pattern (random black dots over white/bright background) over the region of interest of the specimen where the displacements/strains are to be known. Fig. 3 depicts a specimen with typical speckle pattern for LH hole configuration.The process of DIC involves capturing of a reference image at no load and then capturing the images during the loading at specified time intervals during the test. Each image is then processed to obtain the displacements/strains of the speckled region with respect to the reference image.The recording and processing of the images was done using Vic-Snap and Vic-2D software. The images were captured using a PointGrey-GS3-U3-41C6M sensor with Tokina-AT-XPRO MACRO-100F2.8D lens.

3. Results and discussions

3.1. Effect of hole arrangement on load capacity

The strain contours of different specimens were compared to obtain the insight to the effect of arrangement of holes on the behavior of drilled as well as repaired specimens. Fig. 4 illustrates the strain contours developed in different specimens near to their failure. From contours developed for drilled LH specimens, it is observed that in the central region along the length of the specimen between the two holes, the fibers develop very small strains signifying the little contribution they make to the load bearing capacity. Fig. 5 depicts the average ultimate failure load for all the specimen configurations.The study of strain contours developed in repaired LH specimens (LH_SP or LH_DP) reveal further effect of this phenomenon. As observed in LH_SP specimens, the central region of the patch undergoes very little deformations compared to the regions covering the highly stressed hole region. This observation indicates the redundancy of long patch and the over strengthening of the region between the holes by employing the long patches. This signifies the necessity of strengthening of the region nearer the holes along the width by employing a wider patch.Hence the presence of a patch along the longitudinal region between the holes(LH_SP or LH_DP)has a very marginal influence on the load recovery compared to its drilled LH case.

Fig. 1. Schematic of (a) All tested specimens, (b) Ply stacking sequence of Parent Laminate and Patches. (Note- All dimensions are in mm; Notation used- PR: Pristine,SH:Single Hole,LH:Longitudinal Hole,LH_SP:Single Patch Longitudinal Hole,LH_DP:Dual Patch Longitudinal Holes, TH: Transverse Holes, TH_SP: Single Patch Transverse Holes, TH_DP: Dual Patch Transverse Holes).

In case of TH specimens,the presence of two holes transverse to the loading direction significantly decrease the net fibers available to bear the load.The fibers present in the region between two holes experience high stresses. The strain contours reveal the development of high strain zones between the two holes for the drilled specimens. Moreover, for repaired TH specimens (both TH_SP and TH_DP), it is observed that high strains are developed in the patches as well over the region between the two holes revealing a significant transfer of load through the patches. Hence, the presence of the patches provide considerable reinforcement at the repaired zone and results in very high load recovery compared to the TH drilled specimens.This high recovery in case of TH repaired specimens reaffirms the suitability of external patch repair in restoring the structural load bearing capability. However, the comparison between TH and LH repaired specimens revealing lesser recovery in LH repaired specimens compared to drilled,points towards the dependency of the success of patch repair based upon the spread of impact damages (in current study along longitudinal and transverse) with respect to the loading direction.

3.2. Effect of patch type (single vs multiple patches) on load capacity

This investigation was carried to examine the effect of using a large single patch over all damages and multiple (dual) smallpatches bonded over each damage individually. When employing two patches for two holes,the net continuous overlap area of each patch decreases in the vicinity of the hole. The effect of the discontinuity present at the junction between the two smaller patches act as region of higher stress concentration. In case of LH_DP specimens,at higher loads a small gap starts to appear at the region where the two patches meet the two individual patches. A localized high strain at that zone is also observed at the junction as shown in Fig. 4. However, the failure in the repaired LH_DP specimens is initiated primarily under top edge of top patch or the bottom edge of the bottom patch. The post failure pictures of specimens are illustrated in Fig.6.This failure mode is similar to the failure observed in LH_SP case,highlighting no influence of single/multiple patches on repair performance. Furthermore, for TH specimens (in TH_DP) the gap aligns parallel to the loading direction.Thus during the loading,the gap does not open(Fig.4)and its influence become negligible.Hence from these observations,it can be concluded that by using small patches over each hole individually and by using a larger patch spanning over all the damages, a similar post repair behavior is observed.

Fig. 2. Experimental-setup.

Fig. 3. A LH specimen depicting typical speckle pattern.

Fig. 4. Strain contours exhibited by LH and TH specimens.

Fig. 5. Average failure load exhibited by different specimens.

Fig. 6. Typical failure observed in all the specimens.

3.3. Failure modes and tensile response

The study of failure modes aid in obtaining further insight to the behavior exhibited by different specimens. The tested specimens were carefully examined to identify the developed failure modes as depicted in Fig. 6. Fiber breakage is observed in the exterior plies due to the presence of fibers in 0°. Moreover, the fibers in the interior plies are oriented at 45°,which results in the development of significant delamination between exterior 0°and interior 45°plies for the pristine specimens. However, in case of drilled specimens the extent of delamination observed is lesser.The damage in the form of delamination is governed by the rotations induced in 45°degree plies due to the displacement of the specimen. Fig. 7 shows the typical load-displacement curves for all the tested specimens.

It is observed that the pristine specimens undergo a larger displacement before failure,than any of the three drilled cases.The presence of the hole in all the drilled specimens, initiates a significant fiber failure at much lower displacements.This results in the lower relative rotation between 0°and 45°plies in pristine and thus smaller spread of the delamination is observed in the drilled specimens in comparison to the pristine. Moreover, it is observed that the spread of damage in the repaired specimens is also restricted to smaller region, highlighting that the failure of fibers beneath the patch is the primary mode of failure in repaired specimens as well. A sudden fiber failure results in failure of adhesive at patch parent interface and results in ultimate failure in the form of patch slippage in all the repaired cases as depicted in Fig.6.An arrest of these failures is likely to restore the higher strength.In case of TH repaired specimens the presence of patch strengthens the critically stressed fibers beneath it.Thus,a comparison between displacement at failure for TH drilled and TH_DP/SP specimens reveals a higher displacement to failure in repaired TH specimens.This again confirms the participation of the patch in the load transfer in TH configuration and thus resulting in better post repair performance. This phenomenon is not present in LH specimen wherein the displacement to failure for drilled as well as repaired specimens is of similar range.Thus the post repair recovery is very small.Furthermore,in case of LH specimens,two holes are present along the length of the specimen. The damage gets initiated primarily in only one of the hole and the region nearer to the other hole exhibits very little sign of failure.

Fig. 7. Load Displacement plot for (a) Pristine and Drilled Specimens (b) Repaired.

4. Conclusions

In this study the effect of interaction of two holes on tensile behavior of post repaired laminates is investigated.An inspection of failed specimens reveal the influence of hole on the type of ultimate failure induced in the specimens. The patches contribute well to transfer the load through them in TH repair configurations due to their presence over the critically stressed zone. However, the LH specimens have a large area which do not contribute to load transfer, thus having the long patches is redundant in LH hole configuration. The ultimate failure is restricted in the zone of the holes in case of drilled and repaired specimens. Furthermore, it is observed that the difference on tensile behavior due to employing a single large patch and two small patches over each hole is very marginal.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgements

The authors would like to acknowledge the financial support by the Council of Scientific & Industrial Research (CSIR)-Research Scheme (22/0809/2019-EMR-II).