Reduced Order Multiscale Modeling Of Brittle Composites

Research Sponsor:

Air Force Research Laboratory. Program Director: Dr. Stephen Clay.

Research Goal and Objectives

Goal:

Prediction of damage accumulation and failure in composite structures subjected to monotonic and fatigue loadings.

Specific Objectives:

Devise reduced order spatial multiscale methodologies for large scale simulation of failure in brittle composite structures.

Devise temporal multiscale methodologies for the efficient simulation of failure in brittle composites undergoing high cycle fatigue loading.

Apply the new methodologies to modeling carbon fiber reinforced polymers undergoing both monotonic and fatigue loadings.

Research Impact:

Development of a tool to aid composite aircraft design and maintenance.

Tool provides aircraft designers the capability to better understand the failure of composites within large-scale aerospace structures allowing better utilization of the material for higher performance designs.

Enables aircraft maintainers additional information when dealing inducing events on composite aircraft to aid in deciding both the nature and severity of the damage event and make informed decisions concerning aircraft repair.

Provides fundamental understanding composite failure in carbon fiber reinforced polymers subjected to both monotonic and fatigue loadings including information on the damage modes, the order of failure in composites with multiple plies, and the interplay between monotonic and fatigue failure modes.

Capability Development:

The ability to determine the location and type of composite damage in a structural analysis.

The ability to model the damage effects of fatigue loading in carbon fiber reinforced polymers.

Background and Motivation

The Air Force is actively using composite materials in high performance aerospace structures.

A lack efficient and accurate structural life prediction tools for composite materials lead designers to less than optimal designs.

Stress life curves are ineffective for situations where an aircraft has undergone damage such as an impact event.

The complexity of composite materials prevent the straight forward application of current technologies such as metal models.

Technical Challenges:

Gaining the capability to predict the damage growth that occurs at the microscale during structural simulation involving both monotonic and fatigue loadings → Efficient, accurate multiscale methods.

Gaining the capability to account for damage growth over the entire life of a composite aerospace structure → Multitemporal prediction methods.

Designing and implementing the novel algorithms on high performance computing platforms to allow composite life prediction for large scale structural simulations with many elements.

Research Approach

Formulate and implement a combined multiple spatial scale and temporal scale computational framework for modeling damage progression in composite structures undergoing monotonic and fatigue loadings.

Devise a reduced order computational model for modeling damage progression at the microstructural scale for carbonfiber reinforced polymers including a multiplicity of damage modes including transverse matrix cracking, matrix/fiber debonding, delamination, and fiber fracture at a fraction of the cost of a fully resolved microstructural model.

Investigate the damage accumulation response of carbon fiber reinforced polymer composites by undertaking an extensive experimental program involving non-destructive investigative techniques including acoustic emission and x-ray imaging on IM7/977-3 composite specimens.

Eigendeformation-based Reduced Order Homogenization

Reduced Order Microstructure Model:

Symmetric eigendeformation-based reduced order modeling (sEHM) will be employed.

The various composite failure modes such as fiber fracture, transverse matrix cracking, delamination, and matrix/fiber debonding are incorporated naturally using the method of failure paths within the sEHM framework.

Multiple Spatio-Temporal Scale Modeling of Composites

The fixed point temporal homogenization method will be employed and advanced to allow multiscale life prediction in large-scale composite structures subjected to fatigue loadings.

The fixed point temporal homogenization method separates the original boundary value problem into four coupled problems: the macrochronological-macroscopic problem, the macrochronological-microscopic problem, the microchronological-macroscopic problem, and the microchronological-microscopic problem.

Adaptive time stepping metrics are employed to gain high computational efficiency while maintaining good accuracy in predicting fatigue damage progression.