Chapter 14
Time Temperature Superposition Shift Factors for Fabric
Composites
Brian T. Werner and Kevin Nelson
Abstract Physical aging of polymers is a thermodynamic phenomenon that occurs in the glassy regime. Upon cooling, the
thermal contraction is restricted by a lack of adequate free volume within the polymer structure. This leaves the polymer in a
state of thermodynamic non-equilibrium which relieves itself over long timescales. Time temperature superposition is
typically used to accelerate this aging process to achieve validation of properties over the service life of the material. The
shift factors determine the degree to which the material time is accelerated in an isothermal environment at elevated
temperature. This is typically achieved with dynamic mechanical thermal analysis (DMTA). This method works well for
neat polymers but fiber reinforced polymer composites (FRPC) have significantly higher stiffnesses and typical DMTA testing
is limited to under 20 N of force. Due to the large unit cell for a fabric composite and geometrical limitations in the thickness of
a ply, a higher force method would be more useful. In this study, an electrodynamic test frame was used to determine the shift
factors for a glass fiber reinforced polymer (GFRP) composite which has a thermoset matrix. The goal is to determine whether
the shift factors differ for different orientations of the composite. For an orthotropic material, directional dependent shift
factors would increase material model complexity significantly.
14.1 Introduction
A viscoelastic material model for fiber reinforced polymer composites would be valuable not only for determining the lifetime
of a structural component but also for optimizing any processing steps the structure must undergo. On the large time span,
accelerated aging techniques, such as isothermal holds at elevated temperatures, can help evaluate the effect of physical aging
over the projected life of a part without waiting for the materials to age naturally to a full lifetime. This procedure has been
used successfully to produce empirically driven models for aged strength predictions of composite materials based on creep
data [1]. The goal of this study is to similarly investigate the time temperature superposition principal (TTSP) with respect to
fabric composites. Instead of relying on creep data, an approach like that used on neat polymers is attempted. Initial attempts at
formulating a linear viscoelastic model for orthotropic materials has made certain assumptions, one of the largest being that
the horizontal shift factors are not directionally dependent. These shift factors are typically determined by use of isothermal
frequency sweeps at various temperatures from above the glass transition temperature, T g, to significantly below T g and then
stitching the resulting dynamic moduli or tan delta results to encompass a continuous curve over an expanded frequency
range. The degree to which these isothermal frequency sweeps are shifted is referred to as the horizontal shift factor. This type
of study has been difficult to accomplish on fiber reinforced composite materials as they are significantly stiffer than neat
polymers and the force limit for dynamic mechanical thermal analyzers is quite limited (<20 N). While the specimen area can
be reduced for unidirectional composites, fabric composites have a much larger unit cell and even the smallest fabric specimen
can be too large for DMTAs to produce significant strains. Researchers have attempted to get around this limitation by
focusing on bending specimens because much larger displacements can be created using small loads than pure tension.
However, this does not produce a favorable stress state as the coupon has a variable stress state and strain rate through the
thickness. A pure tension approach could theoretically produce cleaner data for model calibration. This issue may be solved
with the advent of electrodynamic load frames that can produce significant loads at higher frequencies. The frame used in this
study is an Instron Electropuls E10000 which has a load capacity of 10 kN and can load at frequencies up to ~50 Hz. With this
control a DMTA type test can be performed on typical tensile coupon geometries.
B. T. Werner (*) · K. Nelson
Mechanics of Materials Department, Sandia National Laboratories, Livermore, CA, USA
e-mail: btwerne@sandia.gov; knelso@sandia.gov
© The Society for Experimental Mechanics, Inc. 2021
R. P. Singh, V. Chalivendra (eds.), Mechanics of Composite, Hybrid and Multifunctional Materials, Volume 6,
Conference Proceedings of the Society for Experimental Mechanics Series, https://doi.org/10.1007/978-3-030-59868-6_14
95
Time Temperature Superposition Shift Factors for Fabric
Composites
Brian T. Werner and Kevin Nelson
Abstract Physical aging of polymers is a thermodynamic phenomenon that occurs in the glassy regime. Upon cooling, the
thermal contraction is restricted by a lack of adequate free volume within the polymer structure. This leaves the polymer in a
state of thermodynamic non-equilibrium which relieves itself over long timescales. Time temperature superposition is
typically used to accelerate this aging process to achieve validation of properties over the service life of the material. The
shift factors determine the degree to which the material time is accelerated in an isothermal environment at elevated
temperature. This is typically achieved with dynamic mechanical thermal analysis (DMTA). This method works well for
neat polymers but fiber reinforced polymer composites (FRPC) have significantly higher stiffnesses and typical DMTA testing
is limited to under 20 N of force. Due to the large unit cell for a fabric composite and geometrical limitations in the thickness of
a ply, a higher force method would be more useful. In this study, an electrodynamic test frame was used to determine the shift
factors for a glass fiber reinforced polymer (GFRP) composite which has a thermoset matrix. The goal is to determine whether
the shift factors differ for different orientations of the composite. For an orthotropic material, directional dependent shift
factors would increase material model complexity significantly.
14.1 Introduction
A viscoelastic material model for fiber reinforced polymer composites would be valuable not only for determining the lifetime
of a structural component but also for optimizing any processing steps the structure must undergo. On the large time span,
accelerated aging techniques, such as isothermal holds at elevated temperatures, can help evaluate the effect of physical aging
over the projected life of a part without waiting for the materials to age naturally to a full lifetime. This procedure has been
used successfully to produce empirically driven models for aged strength predictions of composite materials based on creep
data [1]. The goal of this study is to similarly investigate the time temperature superposition principal (TTSP) with respect to
fabric composites. Instead of relying on creep data, an approach like that used on neat polymers is attempted. Initial attempts at
formulating a linear viscoelastic model for orthotropic materials has made certain assumptions, one of the largest being that
the horizontal shift factors are not directionally dependent. These shift factors are typically determined by use of isothermal
frequency sweeps at various temperatures from above the glass transition temperature, T g, to significantly below T g and then
stitching the resulting dynamic moduli or tan delta results to encompass a continuous curve over an expanded frequency
range. The degree to which these isothermal frequency sweeps are shifted is referred to as the horizontal shift factor. This type
of study has been difficult to accomplish on fiber reinforced composite materials as they are significantly stiffer than neat
polymers and the force limit for dynamic mechanical thermal analyzers is quite limited (<20 N). While the specimen area can
be reduced for unidirectional composites, fabric composites have a much larger unit cell and even the smallest fabric specimen
can be too large for DMTAs to produce significant strains. Researchers have attempted to get around this limitation by
focusing on bending specimens because much larger displacements can be created using small loads than pure tension.
However, this does not produce a favorable stress state as the coupon has a variable stress state and strain rate through the
thickness. A pure tension approach could theoretically produce cleaner data for model calibration. This issue may be solved
with the advent of electrodynamic load frames that can produce significant loads at higher frequencies. The frame used in this
study is an Instron Electropuls E10000 which has a load capacity of 10 kN and can load at frequencies up to ~50 Hz. With this
control a DMTA type test can be performed on typical tensile coupon geometries.
B. T. Werner (*) · K. Nelson
Mechanics of Materials Department, Sandia National Laboratories, Livermore, CA, USA
e-mail: btwerne@sandia.gov; knelso@sandia.gov
© The Society for Experimental Mechanics, Inc. 2021
R. P. Singh, V. Chalivendra (eds.), Mechanics of Composite, Hybrid and Multifunctional Materials, Volume 6,
Conference Proceedings of the Society for Experimental Mechanics Series, https://doi.org/10.1007/978-3-030-59868-6_14
95