The incremental hole drilling method is widely used to determine residual stresses in materials. For fiber-reinforced composites, the drilling-induced heat can create measurement errors that are not taken into account in the calculations. Even if the drilling parameters (cutting and feed speeds, type of cutter, etc.) are optimized to provide correct machining results and more reliable residual stress measurement, thermal effects remain relatively poorly understood and analyzed for fiber-reinforced composites. The purpose of this work is to investigate the effects of drilling-induced heat on the local chemical properties of unidirectional carbon/epoxy laminates that have undergone different curing cycles and to discuss the consequences on residual stresses. First, the thermal field generated by drilling was characterized using an infra-red camera. The measurements showed that relatively high temperatures were reached depending on the increment depth. Then, Modulated Differential Scanning Calorimetry before and after drilling were performed to investigate the effects of the produced heat on the chemical properties. The results showed a local increase in the degree of cure and the glass transition temperature (Tg) of the matrix. It is assumed that these mechanisms are responsible for the apparent residual stresses obtained when large increment depths are used during the process. The main conclusion of this study is that the heat generated during the drilling step of the incremental hole drilling method causes a progression of the cure reaction in the vicinity of the hole, leading to Tg-dependent residual stresses which are higher for low-Tg composites.
Research Article
Measurement of residual stresses in fiber-reinforced composites by the incremental hole drilling method: effect of the drilling-induced heating.
https://doi.org/10.21203/rs.3.rs-6734213/v1
This work is licensed under a CC BY 4.0 License
Journal Publication
published 25 Nov, 2025
You are reading this latest preprint version
Residual stresses
Incremental Hole Drilling Method
fiber reinforced composites
thermal effects
degree of cure
Residual stresses are locked-in stresses in the material that can be very detrimental to performance depending on usage requirements [1-7]. They build up mainly during the manufacturing process [8-10] but can also arise during the lifetime of the parts [11] or be voluntarily generated using techniques such as shot peening [12]. Precise determination of residual stresses and a good understanding of their origins are necessary to optimize the properties of materials and more reliably predict their behavior by enriching existing analytical and numerical models or developing more adapted ones. There are several experimental methods for measuring residual stresses in a wide range of material types. This includes non-destructive techniques such as the curvature method [13] and the X-ray diffraction method [14], destructive techniques such as the layer removal method [15] and semi-destructive methods such as the incremental hole drilling [16-17] and the compliance method [18]. The incremental hole drilling (IHD) is one of the most widely used methods for measuring residual stresses in materials [19]. This method consists of drilling a small hole incrementally through the thickness of the material and measuring the relaxation strains for each increment. These strains are then used to calculate the residual stresses.
One of the long-standing problems with the IHD method is the effect of the drilling itself on the measured data. For this reason, a number of studies have been conducted in this direction mainly on metallic materials [20-23]. To the best of our knowledge, there has been limited research on this topic concerning composite materials. One can cite the work of Yuksel et al. [24] which investigated the effect of drilling on the residual stresses in glass fiber/polyester composite samples. They heated the samples above their glass transition temperature to allow stress relaxation and obtain specimens considered as “stress-free”. They performed IHD with digital image correlation and found that the drilling affected a zone of approximatively 2 mm around the hole. Nobre et al. [25] used a different approach to estimate drilling-induced stresses consisting of applying a known load on the sample to eliminate the effect of initial residual stresses, and then performing IHD. They performed the same test on the same sample by modifying the magnitude of the applied load. The difference between the recorded strains in the first and second tests is due to the difference between the applied loads. Nobre et al. simulated the whole process by the finite element approach and concluded that the difference between experimental and numerical results in terms of strains is due to the effects of drilling.
The studies presented above focus on the mechanical aspect of the drilling and do not take into account the thermal effects due to the process. This is particularly important for composite materials whose thermal, chemical and mechanical properties can be highly interdependent. Considering a relaxation time after drilling allows the material to cool, but does not avoid the irreversible mechanisms induced by heat. In this study, two different increment depths were chosen proportional to the thickness of the layers. The influence of the thermal effects of drilling on the local chemical properties of the matrix is investigated. The consequences of these modifications on residual stresses are discussed.
2.1 Materials:
The materials studied in this work were unidirectional carbon/epoxy composite plates made by filament winding according to the international standard ISO 1268-5 [26]. Two plates of dimensions 100 x 100 mm, with different curing cycles were considered (Table1). The curing cycle of sample 2 is recommended by the manufacturer, and that of sample 1 was chosen to have a lower Tg. They were composed of 12 layers with a thickness of approximately 0.33 mm each. T700SC-24000-50C carbon fibers were used with Araldite 1564 epoxy resin. Aradur 3474 hardener was used as a curing agent. The fiber content of the composites was 70% by mass and 60% by volume. In addition, raw matrix plates were manufactured by molding following the same curing cycles as the composites. 2017A aluminum plate was also considered as reference material for the study of thermal effects on the residual stresses.
Table 1 : Curing cycles of the samples
|
|
Curing cycle |
|
Sample 1 |
10 hours at 70°C |
|
Sample 2 |
1 hour at 80°C then 4 hours at 120°C |
2.2 Methods:
As indicated in the introduction, the IHD method consists of drilling a small hole incrementally through the thickness of the material and measuring the relaxation strains for each increment. For composite materials, these strains are converted to residual stresses using the calibration coefficient matrix (equation 1):
[ε] represents the relaxation strains that are experimentally measured by IHD. [C] is the calibration coefficient matrix that is numerically calculated by the finite element method and used to convert the relaxation strains to residual stresses [19, 27]. The calculation of the residual stresses taking the incremental approach into account is given in equation 2 (Fig. 1) and a more complete description is exposed in [27].
The composite laminates were fixed on the sample holder with 4 glue points on their lower surfaces away from the hole location (more than 50 mm). The edges of the laminates were free. EA-06-062RE-120 strain gage rosettes, designed by Micro-Measurements, were glued to the samples surfaces with gage 1 aligned with the fiber direction. A National Instruments NI 9235 quarter-bridge strain gage module was used to perform the acquisition. A solid carbide milling cutter with a diameter of 2 mm and double-edged face gearing, specially designed for fiber-reinforced composites, was used. In order to study the thermal effects on residual stresses, two increment depths were considered: a large increment depth equal to 0.33 mm (1 increment per ply) and a small increment depth equal to 0.055 mm (6 increments per ply). In addition, IHD was performed on the 2017A aluminum sample, with both increment depths, to compare the influence of the increment depths to that observed on the studied composites (the calculation method has been adapted to an isotropic material [28]).
The spindle trajectories were programmed in such a way that a complete drilling cycle can be described in 5 steps.
- Rapid approach of the milling cutter: from the start position, the milling cutter descends with a speed of 1 mm/s until it reaches 85% of the current depth of the hole (depth obtained after the previous increment).
- Slow approach of the milling cutter: from 85% of the current depth, the feed speed of the milling cutter is slowed down to 0.01 mm/s.
- Drilling step: the increment is drilled with a feed speed of 0.01 mm/s.
- Rapid withdrawal: after the drilling, the milling cutter gets back to the start position with a speed of 1 mm/s.
- Relaxation: for the composite samples, a relaxation time of 90 seconds is observed for the small increment depth and 450 seconds for the large increment depth. For the aluminum sample, a relaxation time of 90 seconds is observed for both increment depths.
The rotation speed and the feed rate used are based on the work of Sicot et al.[29] . All the experimental parameters and conditions are summarized in table 2.
Table 2: Experimental parameters of the IHD tests
|
Test name |
Increment depths |
Rotation speed |
Feed rate |
Ambient temperature |
Ambient humidity |
Test 0.33 Test 0.055 |
0.33 mm 0.055 mm |
5000 rpm |
0.01 mm/s |
22 ± 2 °C |
46 ± 2% |
An infrared camera (ThermaCam SC3000 designed by FLIR) was used to characterize the thermal field during and after the drilling. The thermal tests were performed at an ambient temperature of 22 ± 2 °C, without convective heat flux and isolated from any light source. ThermaCAM Researcher Professional 2.8 SR-2 software was used to post-process the results.
To better investigate the effects of drilling-induced mechanisms on the chemical properties of the samples, modulated Differential Scanning Calorimetry (DSC) tests were performed. A Discovery DSC 25 instrument designed by TA instruments was used with a ramp of 3° /min, a maximum temperature of 250 °C and a modulated temperature of 1° for 60 seconds. Two types of specimens were manually extracted from the raw matrix plates: a cold-extracted specimen and a specimen extracted after drilling on the edge of the hole.
The acquisition and analysis of experimental data are similar for both samples and tests. They are described in this paragraph applied to test 0.33 of sample 1. Fig. 2a presents the raw acquisition data obtained for test 0.33 of sample 1 (table 1 and 2). These data correspond to the strains measured by the gages of the rosette during the incremental hole drilling process.
Each increment cycle is composed of a drilling and a relaxation step (Fig. 2b). The data considered in IHD are the strains at the end of the relaxation time. After the drilling, a redistribution of the local stresses takes place because of the material removal. The relaxation step allows the material to cool down and this redistribution to take place. The duration of the relaxation time depends on the type of material and the induced heat. For test 0.33 of sample 1, at last 360 seconds are necessary to allow the strains to stabilize (Fig. 3a). Using an insufficient relaxation time leads to an overestimation (in absolute terms) of the residual stresses (Fig. 3b).
Fig. 4 shows the thermal field evolution during a complete increment drilling cycle for test 0.33 of sample 1. For the same test applied to sample 2, thermal fields are similar. The results obtained show that during drilling, temperatures reached 55 ± 2 °C at gage 3 location, 60 ± 2 °C at gage 2 location, 80 ± 2 °C at gage 1 location and up to 136 ± 2 °C in the hole (Fig. 4b). Fig. 4a corresponds to the thermal field at the end of the relaxation time of the previous increment. Fig. 4b corresponds to the maximum temperature reached during the drilling of the increment. Fig. 4c shows the temperature during the relaxation time, after the withdrawal of the milling cutter. This thermal field has a principal direction (elliptical shape on the sample surface) due to the thermal conductivity of the fibers. Fig. 4d corresponds to the thermal field at the end of the current relaxation time.
A similar evolution of the thermal field was obtained for test 0.055 for the two plates. However, due to the smaller increment depth, the measured drilling-induced temperatures were about 30% lower (98 ± 2 °C at the hole location vs 136 ± 2°C for test 0.33). These results show that a significant increase in temperature occurs locally during the drilling step, with the heat flow propagating preferentially along the fibers. The question is whether these high temperatures have non-reversible effects on the local properties of the matrix.
Modulated DSC tests were performed on the raw matrix plates corresponding to sample 1 and 2 in order to assess the influence of drilling on the local glass transition temperature of the matrix. A typical DSC acquisition is shown in Fig. 5. All results are given in Table 3.
Table 3: Summary of modulated DSC test results
|
|
|
Glass transition temperature (°C) |
|
Cold extracted |
Sample 1 |
79 |
|
Sample 2 |
103 |
|
|
Extracted after drilling |
Test 0.33 of sample 1 |
99 |
|
Test 0.055 of sample 1 |
83 |
|
|
Test 0.33 of sample 2 |
116 |
|
|
Test 0.055 of sample 2 |
113 |
The results of the cold-extracted specimens were compared to those of the specimens extracted after drilling (table 3). The glass transition temperatures are higher for the specimens extracted after drilling. For test 0.33, an increase of 20°C is obtained for sample 1 and 13°C for sample 2. As expected, the increase in the glass transition temperature was lower for test 0.055 because of the lower increment depth which generates less heating. An increase in the glass transition temperature is indicative of an increase in the degree of cross-linking of the epoxy resin [30].Therefore, drilling causes the cure reaction to progress, modifying the degree of cross-linking and possibly influencing local residual stresses.
The results obtained by the IHD method are valid only for a depth approximately equal to the radius of the hole [31]. This is due to the fact that the gages are less sensitive to deep stresses. Here, a 2 mm diameter cutter is used, so the results are presented on a depth of 1 mm (Fig. 6). The tests were conducted on the composite samples (Fig. 6a) and compared to the aluminum sample (Fig. 6b).
For the composite samples, the results show that the residual stresses are very low for test 0.055 (Fig. 6a). For test 0.33, the residual stresses are significantly higher for sample 1 than sample 2 (in absolute terms). For the aluminum sample (Fig. 6b), the residual stresses are nearly zero and constant for both tests. The increment depth seems to influence the residual stresses of the composite material and not those of aluminum.
Residual stresses are expected to be low in thin unidirectional composite laminates [29] since two of the main mechanisms that induce them do not occur: the high temperature gradient through the thickness during the cooling stage of the manufacture and the mismatch in thermal strain between successive layers with different fiber orientations [10]. Results obtained using smalls increments (test 0.055 of Fig. 6a) are consistent with those expected. The compressive residual stresses obtained for test 0.33 of sample 1 (Fig. 6a) are believed to be apparent stresses induced during the drilling step due to the large increment depth (0.33 mm). In fact, deep increments cause significant heating (Fig. 4) and a change in the local chemical properties of the composite: increase in glass transition temperature (Table 3) and cross-linking [30]. These drilling-induced mechanisms are believed to be the source of the apparent residual stresses. This assumption is supported by the result obtained for test 0.33 of sample 2 and that obtained on the aluminum sample. In fact, the described drilling-induced mechanisms are less important for sample 2 because it is more polymerized. Therefore, compared to sample 1, less apparent residual stresses were found for the large increment depth (test 0.33 of sample 2, Fig. 6a). In addition, for the aluminum sample, for which the described drilling-induced mechanisms do not occur, no apparent residual stresses were found for the large increment depth. Results of test 0.33 were similar to those of test 0.055 (Fig. 6b).
These findings clearly show that, when using the IHD for polymer-based materials cured at low temperatures (low glass transition temperature), it is particularly important to consider small increment depth to limit the heating induced by drilling.
In this paper, the thermal effects of the drilling step of the incremental hole drilling (IHD) method on the local properties of unidirectional carbon/epoxy composite laminates were studied. Investigations were carried out on the influence of the drilling-induced mechanisms on residual stresses.
Two composite samples were studied: a sample cured at 70°C for 10 hours (sample 1) and another one cured at 80°C for 1 hour then at 120°C at 4 hours (sample 2). A test with a large increment depth (0.33 mm) and another with a smaller increment depth (0.055 mm) were performed on each sample. Both tests were also conducted on a 2017A aluminum sample which was considered as the reference for thermal effects. From the obtained results, the following conclusions can be drawn:
- The relaxation time, after drilling, must be long enough to allow the strains to stabilize and reach their asymptotes. Taking an insufficient relaxation time leads to an overestimation of the residual stresses.
- Drilling-induced temperatures reach high enough values (136±2°C at the hole location for the large increment depth and 98± 2 °C for the small one) to increase the glass transition temperature and cross-linking of the matrix. For the large increment depth, the glass transition temperature (Tg) increases from 79 °C to 99 °C for sample 1 and from 103 °C to 116 °C for sample 2.
- The modification of local material properties due to the use of a large increment depth can distort the results and generate apparent residual stresses. This is particularly true for composite with low Tg (sample 1).
- For metallic samples, large increments do not affect the results. The IHD method being time-consuming, this can help reduce measurement time and the number of simulations (needed to determine the calibration coefficients), but only if the desired spatial resolution allows for it.
The effects of drilling-induced mechanisms on the measurement of residual stresses depend on the type of material, its cure cycle, the drilling parameters and the orientation of the fibers for composite materials. To obtain reliable results, it is important to minimize the thermal load particularly for polymer-based material with relatively low Tg.
Funding
The authors did not receive financial support from any organization for the submitted work.
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.
Acknowledgment
The authors would like to acknowledge the G. MAGYAR company for providing the material studied in this research work.
- P. P. Parlevliet, H. E. N. Bersee, et A. Beukers, « Residual stresses in thermoplastic composites – a study of the literature. Part III: Effects of thermal residual stresses », Compos. Part Appl. Sci. Manuf., vol. 38, no 6, p. 1581‑1596, juin 2007, doi: 10.1016/j.compositesa.2006.12.005.
- A. Abedian et W. Szyszkowski, « Influence of the free surface on the thermal stresses in unidirectional composites », Compos. Part Appl. Sci. Manuf., vol. 28, no 6, p. 573‑579, janv. 1997, doi: 10.1016/S1359-835X(97)00008-0.
- J. A. Nairn et P. Zoller, « Matrix solidification and the resulting residual thermal stresses in composites », J. Mater. Sci., vol. 20, no 1, p. 355‑367, janv. 1985, doi: 10.1007/BF00555929.
- W. J. Cantwell, P. Davies, et H. H. Kausch, « The effect of cooling rate on deformation and fracture in IM6/PEEK composites », Compos. Struct., vol. 14, no 2, p. 151‑171, janv. 1990, doi: 10.1016/0263-8223(90)90028-D.
- A. Chiocca, F. Frendo, F. Aiello, et L. Bertini, « Influence of residual stresses on the fatigue life of welded joints. Numerical simulation and experimental tests », Int. J. Fatigue, vol. 162, p. 106901, sept. 2022, doi: 10.1016/j.ijfatigue.2022.106901.
- U. Wiklund, J. Gunnars, et S. Hogmark, « Influence of residual stresses on fracture and delamination of thin hard coatings », Wear, vol. 232, no 2, p. 262‑269, oct. 1999, doi: 10.1016/S0043-1648(99)00155-6.
- H. Wang, H. Li, Y. Xu, Y. Lin, H. Li, et J. Tao, « Effects of Thermal Residual Stresses on Tensile and Interlaminar Shear Behaviors of GLARE Laminates », Appl. Compos. Mater., vol. 28, no 3, p. 877‑898, juin 2021, doi: 10.1007/s10443-021-09902-1.
- J. Lange, S. Toll, J.-A. E. Månson, et A. Hult, « Residual stress build-up in thermoset films cured below their ultimate glass transition temperature », Polymer, vol. 38, no 4, p. 809‑815, févr. 1997, doi: 10.1016/S0032-3861(96)00584-8.
- T. A. Bogetti et J. W. Gillespie, « Process-Induced Stress and Deformation in Thick-Section Thermoset Composite Laminates », J. Compos. Mater., vol. 26, no 5, p. 626‑660, mars 1992, doi: 10.1177/002199839202600502.
- P. P. Parlevliet, H. E. N. Bersee, et A. Beukers, « Residual stresses in thermoplastic composites—A study of the literature—Part I: Formation of residual stresses », Compos. Part Appl. Sci. Manuf., vol. 37, no 11, p. 1847‑1857, nov. 2006, doi: 10.1016/j.compositesa.2005.12.025.
- M. N. James, « Residual stress influences on structural reliability », Eng. Fail. Anal., vol. 18, no 8, p. 1909‑1920, déc. 2011, doi: 10.1016/j.engfailanal.2011.06.005.
- H. Liu et al., « Effect of stress shot peening on the residual stress field and microstructure of nanostructured Mg-8Gd-3Y alloy », J. Mater. Res. Technol., vol. 10, p. 74‑83, janv. 2021, doi: 10.1016/j.jmrt.2020.11.085.
- S. R. White et H. T. Hahn, « Process Modeling of Composite Materials: Residual Stress Development during Cure. Part I. Model Formulation », J. Compos. Mater., vol. 26, no 16, p. 2402‑2422, janv. 1992, doi: 10.1177/002199839202601604.
- S. A. Martinez, S. Sathish, M. P. Blodgett, et M. J. Shepard, « Residual stress distribution on surface-treated Ti-6Al-4V by X-ray diffraction », Exp. Mech., vol. 43, no 2, p. 141‑147, juin 2003, doi: 10.1007/BF02410495.
- M. Eijpe et P. Powell, « Residual stress evaluation in composites using a modified layer removal method », Compos. Struct., 1997, doi: 10.1016/S0263-8223(98)80004-4.
- N. J. Rendler et I. Vigness, « Hole-drilling strain-gage method of measuring residual stresses », Exp. Mech., vol. 6, no 12, p. 577‑586, déc. 1966, doi: 10.1007/BF02326825.
- G. S. Schajer et L. Yang, « Residual-stress measurement in orthotropic materials using the hole-drilling method », Exp. Mech., vol. 34, no 4, p. 324‑333, déc. 1994, doi: 10.1007/BF02325147.
- N. Ersoy et O. Vardar, « Measurement of Residual Stresses in Layered Composites by Compliance Method », J. Compos. Mater., vol. 34, no 7, p. 575‑598, avr. 2000, doi: 10.1177/002199830003400703.
- S. Akbari, F. Taheri-Behrooz, et M. M. Shokrieh, « Characterization of residual stresses in a thin-walled filament wound carbon/epoxy ring using incremental hole drilling method », Compos. Sci. Technol., vol. 94, p. 8‑15, avr. 2014, doi: 10.1016/j.compscitech.2014.01.008.
- M. T. Flaman, « Brief investigation of induced drilling stresses in the center-hole method of residual-stress measurement », Exp. Mech., vol. 22, no 1, p. 26‑30, janv. 1982, doi: 10.1007/BF02325700.
- M. T. Flaman et J. A. Herring, « Comparison of Four Hole-Producing Techniques for the Center-Hole Residual-Stress Measurement Method », Exp. Tech., vol. 9, no 8, p. 30‑32, août 1985, doi: 10.1111/j.1747-1567.1985.tb02036.x.
- C. C. Weng, Y. C. Lin, et C. P. Chou, « A New Approach for Determining the Induced Drilling Stresses in the Hole-Drilling Method of Residual-Stress Measurement », Exp. Tech., vol. 16, no 5, p. 33‑35, 1992, doi: 10.1111/j.1747-1567.1992.tb00704.x.
- A. K. Tiwari, A. Kumar, N. Kumar, et C. Prakash, « Investigation on micro-residual stress distribution near hole using nanoindentation: Effect of drilling speed », Meas. Control, vol. 52, no 9‑10, p. 1252‑1263, nov. 2019, doi: 10.1177/0020294019858107.
- O. Yuksel, I. Baran, N. Ersoy, et R. Akkerman, « Investigation of transverse residual stresses in a thick pultruded composite using digital image correlation with hole drilling », Compos. Struct., vol. 223, p. 110954, sept. 2019, doi: 10.1016/j.compstruct.2019.110954.
- J. P. Nobre et al., « Quantifying the Drilling Effect during the Application of Incremental Hole-Drilling Technique in Laminate Composites », Mater. Sci. Forum, vol. 681, p. 510‑515, 2011, doi: 10.4028/www.scientific.net/MSF.681.510.
- « Standard - Fibre-reinforced plastics - Methods of producing test plates - Part 5: Filament winding ISO 1268-5 », Svenska institutet för standarder, SIS. Consulté le: 28 janvier 2022. [En ligne]. Disponible sur: https://www.sis.se/produkter/gummi-och-plastindustri/armerad-plast/iso12685/
- A. S. Ibrahim Mamane, S. Giljean, M.-J. Pac, et G. L’Hostis, « Optimization of the measurement of residual stresses by the incremental hole drilling method. Part I: Numerical correction of experimental errors by a configurable numerical–experimental coupling », Compos. Struct., vol. 294, p. 115703, août 2022, doi: 10.1016/j.compstruct.2022.115703.
- J.-N. Aoh et C.-S. Wei, « On the Improvement of Calibration Coefficients for Hole-Drilling Integral Method: Part I—Analysis of Calibration Coefficients Obtained by a 3-D FEM Model », J. Eng. Mater. Technol., vol. 124, no 2, p. 250‑258, avr. 2002, doi: 10.1115/1.1416685.
- O. Sicot, X. L. Gong, A. Cherouat, et J. Lu, « Determination of Residual Stress in Composite Laminates Using the Incremental Hole-drilling Method », J. Compos. Mater., vol. 37, no 9, p. 831‑844, mai 2003, doi: 10.1177/002199803031057.
- X. Wan, B. Demir, M. An, T. R. Walsh, et N. Yang, « Thermal conductivities and mechanical properties of epoxy resin as a function of the degree of cross-linking », Int. J. Heat Mass Transf., vol. 180, p. 121821, déc. 2021, doi: 10.1016/j.ijheatmasstransfer.2021.121821.
- G. S. Schajer, « Hole-Drilling Residual Stress Measurements at 75: Origins, Advances, Opportunities », Exp. Mech., vol. 50, no 2, p. 245‑253, févr. 2010, doi: 10.1007/s11340-009-9285-y.
No competing interests reported.