Tensile Analysis of Fabric Reinforced Materials

The analysis of reinforced polymers is one of the most important sources of information for materials properties designers. In this regard, due to their intrinsic properties, the fabrics play an essential role when they are used as reinforcement elements. The composites properties can be designed by alternating various types of fabrics, by modifying the orientation of every reinforcement layer, by modifying the matrix properties or by choosing the matrix. This study regards the tensile behavior of four fabric reinforced composites with four different epoxy resins as matrix. All the materials have the same reinforcement structure but the matrix is, in each case, another epoxy resin and more two classes of materials had been studied one is containing the natural polymerized matrix materials and the other one the materials that had been thermally treated according to technical sheet of each polymer. The tests were done one year after the materials formation.

REINFORCEMENT STRUCTURE The main advantage of laminates is that, based on forming technique, they allow a free distribution of fibers arrangement [41][42][43][44]. To meet the objectives of the study, eight laminates with reinforcement and homogeneous matrix were formed. The first aim of the study is to identify the role of the matrix on the tensile behavior of laminates (as long as all of them have the same reinforcement structure) and the second aim is to identify the influence of the thermal treatment on the tensile behavior of materials (since for the each type of matrix one formed laminate was thermally treated and the other one was not).
In other studies, the behavior of laminates had been analyzed under tensile conditions [26,40,45], as well as the effect of the modification of the epoxy resin, the effect of the type of fabric used in the reinforcement of the layers, but also the influence of the variation of the number of layers of each fabric (carbon, aramid, glass) as well as the influence of changing the fibers orientation in the laminates reinforcement [46][47][48].
The standard tensile samples were extracted from the formed plates at one year after their formation using a high pressure water jet cutting machine fig. 2a. The tensile tests were done on an Instron machine with maximum loading capacity of 100kN. The tests were set for a loading speed of 5mm/minute and the stop condition was set at 50% drop of force or 50mm displacement of grips. Al tests were video recorded (with two video cameras) in order to point out the precise moment of the failure. Also the testing segment of each sample was monitored using thermo-vision Fluke Ti400 video camera in order to identify eventual heating of samples during the tests. More than that, it was ensured the identification of each sample position ( fig. 2b) into the original plate ( fig. 2c) with the aim of explaining possible different behaviors of certain samples (the ones extracted from the edges of original plates or the ones affected of invisible defects such as matrix discontinuities).

Results and discussions
All the samples for a certain material had been tested during one day period (in fact three to four hours) to ensure, practically, the same value of the laboratory air temperature. One video camera was positioned such as to record one face of the sample while the second one was positioned such au to record one of the lateral edges of the sample. The decision to do that was made on the basis of the necessity to correlate various critical points on the loading curves with recorded moments on sample evolution but because of the low resolutions of the two cameras the only success was to extract images of the samples at the very moment they failed. The thermos vision camera was laced such as to record both a face and an edge.
In order to realize a the best analysis of the results all the results were statistically analyzed end for each material the averaged curve was realized using three to five partial results, corresponding respectively to three to five tested samples their numerical marks being mentioned on the general curves graphs.
On the right corner of general graphs there are presented two images (a front one and a lateral one) of the samples with the lowest response on the respective series. On another hand, on the graphs presenting the averaged curves there are presented, together with front and lateral images, the images of the highest value of temperature reached by a sample selected by statistical meansas part as averaging process.
In the case of C matrix materials - fig. 3.it is easily to notice that in the case of natural polymerized materials the dispersion data of individual results is larger than the dispersion data of thermally treated materials but the differences are not too significant. It is interesting that the samples with the lowest response for both series are the ones numbered with 8 viz. the ones at the edge of composite plate and because of that perhaps their structure was affected by the high pressure water jet tacking into account the fact the penetration of the plate by the water jet started with these samples. After the 8 samples are detached the composite plate rigidity is lowered and it is less probable to affect the next future samples.
It has to be said that regarding the samples with approximately same behavior the sample numbered as 5 was the one having the highest increase of temperature value for both natural polymerized and thermally treated materials. The difference between the highest (sample 5) and lowest (sample 6) values of temperatures for natural polymerized composite material is of 18.64°C while the other two samples recorded values of temperature around the average of the extremes. On the graphs of C materials is observable the fact that the heating of sample 5 is extended between the two fracture points. On the graph of averaged curve for CT materials (thermally treated) the differences between extreme values of temperature is of 5.74°C (48,49°C for sample 5 and 42.75°C for sample 6). In this case for the sample 1 was recorded a value of 48.47°C (practical the same as for sample 5) and for the last sample the recorded value was 45.09°C.
Analyzing just the thermal aspects it might be said that the thermal treatment contribute to the composite material consolidation leading to better quality interfaces. There is something else namely the fact that the CT materials (selected after statistical analysis) are failing by fracture of the external block of carbon fibers on one face of the sample while the natural polymerized materials are failing by fracture of both blocks of carbon fiber fabric reinforcement. Generally, for the C materials and CT materials the core of the reinforcement is not affected for the case of all averaged samples.
In fig. 4. there are presented the curves for individual samples and the pictures in the right corner corresponds to the samples with the lowest recorded tensile response. In the case of E materials, the sample number 6, and in the case of ET materials, again, the sample number 8. Regarding the poor response of the sample 6 perhaps its behavior is due to some internal defects (lack of polymer between two or more reinforcement sheets) and the fracture of the sample affected both packets of external reinforcement layers with their displacement from the core of reinforcement. The difference of heating temperature values for E materials is of 17.97°C (between sample 1 -and sample 8) the other two values being around the average of extremes. In the case of ET materials (sample 6 and sample 3, respectively) the temperature difference is of 3.31°C but the highest value is reached in the case of sample 4 and is of 51.68°C. From the mechanic point of view it is easily to notice that the dispersion of sample responses is higher in the case of natural polymerized materials than the case of thermally treated material.
Regarding the H matrix materials, the results are shown in fig. 5. As in the case of other presented materials the dispersion of tensile testing results are more dispersed in the case of naturally polymerized material (H materials) and, also as in the already presented results, for these materials the fracture affected both blocks of external layers of reinforcement (the ones made of carbon fiber fabric C240) but in some cases (the ones there are presented as having the lowest response -H matrix sample number 4 and, HT matrix sample number 8the core of the material is also affected. The averaged curves for these materials are made over five samples in the case of H material and over four samples in the case of HT material. The heating effect at the break is present also and it is of highest value of 12.4°C (57.48°C for sample 1 and 45.08°C for sample 5 and sample 6) for the H matrix material and is of highest value of 5.77°C (sample number 7 and sample number 5) for the HT matrix material. The low dispersion of the results denotes that in the case of H matrix materials the matrix-reinforcement interphase could be of better quality. The last two materialsjust because we assumed to present them alphabeticallyare presented in fig. 6. In this case it easily to notice that the dispersion of the naturally polymerized samples did not allowed an average over more than three samples. In the case of L matrix material the lowest response is, again, the one of sample number 8 while in the case of HT matrix material the lowest response was considered the one of sample number 7. For these composite materials it was noticed that the core of the reinforcement is affected by tensile tests (in both cases of naturally polymerized matrix -L and, thermally consolidated matrix -LT). Regarding the increase of temperature value there were recorded a difference of 8.07°C, for the L matrix material and, of 13.02, for the LT matrix material (which seems to be opposite from the other three materials). It has to be mentioned the fact that the highest temperature values reached during these materials testing were 62.20°C for E material, sample number 7 and, 65.09°C for ET material sample number 2.
Regarding the heating of the samples at break it might be assumed that the final value of the temperature is connected both with the quality of the matrix-reinforcement interphase and with the amount of energy released when the carbon fibers are breaking. It is possible to see that the high temperature zones are corresponding mainly to the external reinforcement layers (carbon fibers ones) for all the studied materials. In Table 2 the values of temperature at break are presented for all the tested samples and these values should interpreted as mechanical effects but at this moment the values of specific heat capacity of materials are not available. On dark background are presented the temperature values at break for the samples that were used to build up the averaged curve for each material. The average temperature values had been evaluated both for the entire set of samples and just for the samples with closed mechanical responses and the values are given in Table 3. These values are corresponding not to highest temperature values but for the increase of the temperatures and they had been obtained after the initial temperature value of the sample was removed.
The assumption that there it is possible to connect the increase in temperature value and some mechanical effects induced by tensile loading, fails when analyzing the results presented in Table 3. It is possible to notice that, generally, the average increase in temperature is slightly larger in the case of thermally consolidated materials excepting the L material when the average value of temperature value increase is higher for the naturally polymerized material. Also it is possible to notice that the average temperature value for C and E materials in the case of selected samples (it is about the statistic selection) is higher than in the case of naturally polymerized correspondents and it is reversed for the case of H and L materials. From previous studies it is known that H and L polymers (Epoxy Resin HT and Epoxy Resin L) are more brittle than the other two resins and an explanation could be found on this, meaning that micro-fractures of matrix lead to local heating of materials residing in a higher temperature at the end.
Analyzing the averaged curves in fig. 3. to fig. 6. It is possible to identify some passages of the mechanical behavior of all the samples as in fig. 7. The first passage is very short and almost horizontal and it might correspond to the clamping of the sample between pneumatic grips together with the start of the loading (this first passage is not marked in fig. 7.). The second passage corresponds to the elastic response of the material while the passages three, four and five, could correspond to failures of certain interphases like passage four that could be associated to the detachment of carbon fiber layers from the others from the core. The last passage for each material correspond to the linear response of carbon fibers fabrics out of matrix even in this case it is difficult to analyze all the mechanical loadings because the aspect of the fractures of the carbon fiber reinforcement packets at external layers look mainly as bending induced fractures and, for some samples after fracture they are rotate relatively to the core (which is maintained in loading position by the grips) signalizing that some shear efforts are appearing. In Table 4. the values of all the slopes are presented for each averaged curve and, of course, just the second could be considered elastic modulus. The elastic modulus seems to be higher for the thermally treated materials excepting the H matrix material. For the last three materials (E matrix, H matrix and L matrix) the final passages (the fifth and the sixth) are merged so the same values of correspondent slopes are presented. The final passage shows almost the same value confirming the fact that the final response of the material is a fiber response. The higher value of the elastic modulus is reached for the E matrix materials (de Epiphen resin is the less brittle one).

Conclusions
Eight materials with four different epoxy resins (commercially available) were formed by to analyze the effect of consolidating the matrix on the tensile properties of materials. Each material is reinforced with 35 layers of fiber fabrics with five sheets of glass fibers at the very core, that are protected by aramid fiber fabrics (another five sheets each side) covered by mixed fabrics made of aramid fibers and carbon fibers (another five sheets each side) and at the end (as external layers) carbon fibers fabric (another five sheets each side).
Two plates of 250x340mm were formed for each matrix and one of the plates was thermally consolidated while the other one was not. The mechanical tests were developed one year after the materials formation. The sample were extracted from the original plates by using a high pressure water jet machine and the position of each sample was marked in order to associate eventual mechanical effects with the sample position.
The experimental fixture included two video cameras (one to record a face of the sample and, one to record an edge of the sample) in order to associate the mechanical tests events with the visual effects (breaking mostly). A thermos-vision camera was used to monitor the increase of temperature during tests and the results concern mostly with a release of energy (identifiable by increase of temperature value) at break.
Statistical analysis of partial results leads to the conclusion that only few of the samples can be used to characterize the materials and with these results the averaged curves σ/ε were evaluated. Each time the samples that were used to average are mentioned on graphs together with two imagines of the fractured sample (frontal and lateral) that had presented the highest increase in temperature value. Also on the general graphs of all the samples for a given material two imagines (frontal and lateral) were presented for the sample that showed the lowest response.
The thermal measurement shows that there are differences between the consolidated materials and the one that are naturally polymerized but this values (averaged bot for the entire lot of samples or just for the similar mechanical behavior samples) are not too different and they cannot be used to interpret the mechanical results.
The mechanical behavior of materials, given by the averaged curves, presents for each material six passages. From these passages, the second corresponds to elastic response of the composite material while the sixth one represents the elastic response of the fibers. The highest value of the elastic modulus is reached for Epiphen RE4020-De40202 epoxy system (around 5Gpa) and it corresponds to the less brittle matrix (for both materialsthe one consolidated by thermal treatment and the one that is naturally polymerized).