th  A probabilistic structural analysis of an experimental Composite Crew Module (CCM) of the future spacecrafts was performed using Orion Crew Module design loads obtained from the NASA Johnson Space Center. The purpose was to quantify the probability of failure for different factors of safety in response to uncertainties in the design variables. The results of these reliability calculations provide a quantitative means for selecting appropriate composite materials. An additional purpose was to demonstrate how probabilistic methods provide cost savings by allowing a reliability manager the ability to select proper factors of safety for predetermined risk.

Structural analyses were performed for 13 different load cases and the most critical load condition and the corresponding critical regions of high stresses were identified. A high resolution analysis at ply level was conducted in that region to determine peak stresses and identify potential failures. The peak stress and type of failure were then used in performing the probabilistic structural analyses.

Based on probabilistic structural analysis, the CCM is found to be very safe with ample margin of safety and a low probability of failure. As a further analysis activity, the loads were scaled up and the composite material’s strength was scaled down to assess a bounding scenario. Still, the design is found to be safe.

Probabilistic methods demonstrate how to select materials for the structure based on setting tolerances (cost) and factor of safety for predetermined risk. Without using probabilistic methods, the reliability of the design remains unknown; in contrast to using a Factor of Safety method alone.

Finally, probabilistic methods provide one more important piece of information for key managers when optimizing the use of the critical resources. This information consists of sensitivities of the input variables such as geometry, material properties, etc., on the response variables such as stress. For improving the quality of the end products, managers need to know which resources have high sensitivities, and thus are controlling the structures’ performance or safety.

Services Provided:

  •  Identified critical areas with high stresses.
  • Estimated the risk of failure for operating conditions in space including the uncertainties in geometry, material properties and loading conditions.
  • Established a relationship between a factor of safety and risk.

This slideshow requires JavaScript.

Composite Crew Module of Future Space Vehicles

N&R Engineering conducted a workshop at NASA that demonstrated the advantages of using probabilistic analysis methods. These methods were applied as part of a structural analysis of the Multi Mission Space Exploration (MMSEV) Crew Cabin. This structure whose outer shell is composite is analyzed for the highest principal stress at the worst-case composite ply location from several load cases. Stress and strain results are input into several composite failure criteria using an in-house code to determine failure indices of the composite ply.

As part of the probabilistic analysis, several material properties and loads were perturbed to demonstrate each variable’s influence on the ply stress and ultimately, the failure index of the composite. The resulting stresses or failure index (response variable) from each separate perturbed variable solution were input using the NESSUS code to predict probability density functions of the response variable and sensitivities of the perturbed variables. This method can be a powerful tool in identifying which variables have significant influence on the structure and have been used on past projects to make decisions on material testing to redirect funding. The workshop was intended to demonstrate the hands-on use of the fore-mentioned codes using the MMSEV as a sample problem.

Services Provided:

  • Structural Finite Element Analysis
  • Composite Structural Analysis
  • Sub-Modeling of Critical Composite Area
  • Probabilistic Failure Modes Analysis
  • Developed Code to Translate NASTRAN Results as Inputs to the NESSUS Probabilistic Code

This slideshow requires JavaScript.

> Back to Completed Work