Experimental Investigation on The Performance of Vacuum Infused Composites

Abstract

In this work, the problem of dry area formation during vacuum infusion is studied. In particular the effect of dry area repair by re-infusion on the tensile properties of the composite material are investigated. The work presented a comparison between repaired (re-infused) and unrepaired (Intact) samples produced by VARI. The ultimate tensile strengths (UTS) were measured and the failure modes were observed for the two cases. A methodology using finite element analysis program "PAM-RTM" was presented to prevent the occurrence of dry areas during VARI. A 25 m long wind turbine blade is taken as a case study. The following conclusions can be drawn: 1- Re-infusion has a minor effect on the ultimate tensile strength especially when the parent and repair laminates have almost the same quality. i.e. (fiber volume fraction and void content). The ultimate tensile strengths for uni-axial fabric UNIE1050M50 and the tri-axial fabric ETXL 1200 (-45º /+45º/ 0º) are reduced by 10.3% and 3.9 % respectively for re-infused samples.

2- If the repair has higher quality than the parent laminate, the resultant strength may exceed the baseline strength of the intact laminate. This is the case of tri-axial fabric ETXT 900 (+45º/90º/-45º) and the woven roving fabric WR600 where the performance of the parent laminate was considerably improved by the high quality repair which led to increase in the strength of the re-infused coupons with respect to the intact. A 15.4% and 18.5% increase in the ultimate tensile strength are recorded for ETXT 900 and WR 600 respectively after the re-infusion.

3- For each type of fabric, there is a typical dominant failure mode. For uni-axial fabric UNIE1050M50, delamination is the dominant failure mode. While for both tri-axial fabrics ETXL 1200 and ETXT 900, the dominant failure mode is angled fracture at 45º due to the presence of ±45º plies in the fabrics. For woven roving coupons WR 600, lateral fracture is the dominant failure mode due to the presence of 90º bundles in the fabric. 4- It is noticed for all types of fabrics that failure occurs usually in the laminate which has higher void content and lower fiber volume fraction. For both tri-axial fabrics ETXL 1200 and ETXT 900, failure sometimes occurs at the interface between parent and repair laminates. 5- In re-infused uni-axial samples failure usually stops at the interface between parent and repair laminates, which means that re-infusion may be useful to prevent crack propagation. 6- The main reason for dry area formation during infusion is the improper design for infusion strategy. Thus, numerical simulation should be done to prevent it. 7- The most important input for flow simulation is the permeability of the reinforcement which has to be measured. Therefore, experimental tests were carried out to measure the permeability in x,y and z directions using 1D channel flow method. The permeability results were validated using a numerical model. Small deviations were reported between experimental and numerical model.

8- For the 25 m wind turbine blade, different strategies were simulated using "PAM-RTM" to impregnate the upwind half with resin. It was concluded that seven feeding lines can fill the blade before resin starts to cure so that no dry areas would result. The distance between two successive feed points along the feed line was 4m. The vacuum lines were distributed at the leading and trailing edge of the blade.