Due to the time and resource intensive installation process, traditional electrical strain gauges are often deployed in limited numbers. Because of budgetary or schedule constraints, engineers are often forced to determine critical points throughout the structure where individual strain gauges will be installed. This lack of comprehensive coverage creates unforeseen risks that can translate into catastrophic failure. Ignorance is not bliss when it comes to testing. FOS technology enables engineers to capture significantly more data than they can practically with strain gauges and ensure that potential problematic areas on a structure are not missed.
One application amongst many where FOS has demonstrated its usefulness is in the structural testing of a large-scale sandwich composite cylinder (CTA8.1) at NASA Marshall Space Flight Center, shown below in Figure 1.
Figure 1: (a) CTA8.1 Test Assembly
(b) Finite element model of CTA8.1 Test Assembly.
The testing was completed in 2016 under NASA’s Shell Buckling Knockdown Factor (SBKF) Project. The goal of the project was to improve thin-walled shell buckling design guidelines widely used throughout the aerospace community that have not been updated since the 1960s. For modern launch vehicles, buckling is often the critical failure mode and the current guidelines have proven to be overly conservative in most cases. Under improved design guidelines, structural margins can be confidently reduced while maintaining safety of the structure, yielding significant weight savings. For modern launch vehicles, significant costs savings per launch are realized due to reduced material costs and increased payload capacity.
The composite cylinder (CTA8.1) was instrumented with 16 optic cables (40-foot long), each containing over 1,000 individual fiber optic strain sensors (0.50” gauge length). The 16 fibers were directly connected to a modified version of the Sensuron RTS125+ FOS system. The installation layout, shown in Figure 2 (a), included eight axial (vertical) runs and five circumferential (horizontal) runs. Additionally, 144 individual fiber optic strain rosettes were implemented near the top and bottom interface rings (Figure 2 (b)). Note that the installation is mirrored on both the outer mold line (OML) and inner mold line (IML) surfaces (not obvious in the Figure).
Figure 2: (a) FOS Instrumentation layout on the outer mold line
(b) FOS rosette layout.
The axial and circumferential fiber optic cables (Figure 8 (a)) are installed in order to capture the global strain distribution throughout the cylinder for the subcritical load cases as well as to identify the failure location for the critical load case. The FOS strain gauge rosettes are installed near the top and bottom attachment rings (Figure 8 (b)) to monitor the uniformity of the load introduction onto the composite cylinder. In addition to the 16,000 fiber optic strain sensors, traditional measurements were collected from 256 electric strain gauges, 28 displacement transducers, and 14 digital image correlation (DIC) photogrammetry systems. All 3 technologies used as part of this experiment were complementing one another.
The test article was initially subjected to a series of subcritical axial compression load cases (ranging from 20% to 50% ). At these subcritical load steps, the axial fiber runs confirmed that the axial strains were being distributed essentially uniformly. During the final load case, uniform axial compression was incrementally applied until buckling failure occurred. As the applied load approached , deformations throughout the structure began to produce a non-uniform internal load distribution due to stiffness changes.
As thin-walled shells are loaded critically in compression, the amount of bending present is a useful parameter to monitor as it is indicative of the amount of radial deformation occurring in each panel, leading to buckling. The FOS sensors were purposely installed on the OML and IML surfaces to characterize the bending strain in each panel. In the Figure below, the bending strain distribution throughout the cylinder is shown just prior to the failure.
Figure 3: (a) Distributed bending-strain measurements just prior to failure
(b) Bending-strain intensity color scale.
The bending strains shown in Figure 9 are calculated as one-half the difference between the IML and OML strain measurements. Since the FOS system was operated at a 0.50” spatial discretization, the IML and OML strain measurements were aligned within 0.50”. As shown in the Figure, the maximum bending strain occurred at approximately 10 inches above the mid-height of the cylinder at the 45° circumferential position. Pre-test buckling predictions often deviate from actual buckling, making it difficult to identify critical points for traditional strain gauges to be installed. The large spatial coverage provided by the 16,000 FOS sensors greatly increased the probability of capturing bending strain from the critical panel and location.
The distributed data from the FOS measurements provided a high resolution strain map of the axial and circumferential strain occurring throughout the structure. Additionally, the mirrored FOS installation made it possible to monitor the bending strain occurring throughout each panel as the applied load approached Pcr.