Modeling Gas Dynamic Effects in Shock-Tubes for Reaction Kinetics Measurements

Kevin Grogan, Mengqing Wang, Matthias Ihme53rd 2015 · DOI

TL;DR

This paper introduces STANSHOCK, a computationally efficient, low-order model that accurately captures the non-ideal gas dynamic effects (like boundary layer growth) in shock tubes. This model is crucial for improving the accuracy of chemical kinetic measurements derived from shock tube experiments, which are often skewed by these effects.

What Problem Does This Paper Address?

Shock tubes are vital experimental tools for studying chemical reactions, particularly for determining ignition delay times and reaction rates. The ideal scenario in a shock tube involves a stationary and uniform gas mixture behind a reflected shock wave. However, real-world shock tubes are subject to non-ideal phenomena, primarily viscous and heat transfer effects at the walls, leading to boundary layer growth. This boundary layer can cause the incident shock wave to attenuate (lose strength) and, more importantly, leads to a temporal temperature rise behind the reflected shock. This temperature rise can significantly alter the reaction kinetics, leading to inaccurate measurements of rate coefficients and ignition delays. Existing models often address these effects post-experimentally or are computationally expensive. This paper aims to develop a model that can simulate these gas dynamic effects within the shock tube environment, allowing for better experimental design and more accurate kinetic data interpretation.

Methods

The core of this work is the development of a one-dimensional, low-order model called STANSHOCK. This model is derived by cross-sectionally averaging the axisymmetric, reactive Navier-Stokes equations. To account for the non-ideal effects, additional "sink terms" are incorporated into the momentum and energy equations to represent losses due to heat transfer and viscous effects from the boundary layer.

Key aspects of the method include:

  • Quasi-1D Governing Equations: The model uses simplified, averaged equations that capture the essential physics without the full complexity of 3D simulations.
  • Boundary Layer Modeling: Viscous shear stress at the wall is modeled using the Karman-Nikuradse skin-friction correlation, and heat loss is modeled using a Stanton number correlation for turbulent flows. These correlations are standard in fluid dynamics.
  • Numerical Scheme: The governing equations are solved using an adaptive mesh refinement (AMR) method implemented in the AMROC framework. Spatial derivatives are discretized using a second-order finite-volume scheme, and a forward Euler method is used for temporal integration. A Minmod limiter is employed for shock capturing.
  • Comparison with URANS: The STANSHOCK model's performance is validated against more computationally expensive unsteady Reynolds-averaged Navier-Stokes (URANS) calculations using OpenFOAM, employing both k-ω and Spalart-Allmaras turbulence models.

Datasets

Not discussed in the paper. The paper uses experimental data from Pang et al. (Ref. 15) for validation of the pressure rise and Petersen's experimental relation (Ref. 19) for shock attenuation. A simulated experiment is conducted using a known chemical mechanism (Hong et al., Ref. 21) for hydrogen peroxide decomposition.

Key Insights and Ideas

  • Boundary Layer Effects are Significant: The study clearly demonstrates that boundary layer effects in shock tubes are not negligible and can lead to substantial errors (e.g., a 23% discrepancy in activation energy) in kinetic measurements if not accounted for.
  • STANSHOCK Offers a Computational Advantage: The low-order STANSHOCK model provides results comparable to expensive URANS simulations but requires orders of magnitude less computational resources, making it ideal for parametric studies and experimental design.
  • Isentropic Modeling as a Viable Approximation: For certain experimental conditions, particularly at higher temperatures, modeling the shock tube as an isentropic reactor using a measured pressure trace can effectively account for boundary layer effects, offering a simpler alternative to full CFD.
  • UV Reactor Limitations: Simple Unsteady (UV) reactor models, which assume adiabatic and isochoric conditions, perform poorly when boundary layer effects are significant, often over-predicting reaction rates.

Key Results

  • STANSHOCK accurately replicates experimental observations of temporal pressure rise behind the reflected shock and incident shock wave attenuation.
  • Compared to URANS models, STANSHOCK achieves similar accuracy in predicting pressure rise but with approximately 1/100th of the computational cost.
  • The model predicted a shock attenuation of approximately 0.6%, which is close to empirical predictions and within experimental uncertainty.
  • In a simulated experiment, UV reactor modeling in the presence of boundary layer effects over-predicted the rate coefficient by a factor of 2.1 at the lowest temperature, leading to a 23% decrease in the calculated activation energy compared to the actual kinetic rate.
  • Isentropic modeling, when using a known pressure trace, was found to accurately reproduce the actual rate coefficient for the simulated experiment, suggesting its viability for specific conditions.

Limitations and Future Work

  • General Correlations: The authors acknowledge that the use of general skin-friction and Stanton number correlations might limit the precision of STANSHOCK. Tailoring these correlations to shock-tube specific gas dynamics could improve accuracy.
  • Heat Loss Modeling: Heat loss to the wall in the test section is not explicitly modeled in STANSHOCK, which could contribute to over-predictions in pressure.
  • Diaphragm Rupture: The model assumes an ideal Riemann problem initialization, neglecting the non-ideal diaphragm rupture, which can affect shock wave attenuation.
  • Future Work: The authors suggest incorporating effects that counteract non-ideal gas dynamics, such as driver inserts, into future modeling efforts. They also propose using STANSHOCK to investigate detonation wave formation and simulate tailored experiments.

Summary updated

BibTeX
@inproceedings{grogan2015modeling,
  author = {Kevin Grogan and Mengqing Wang and Matthias Ihme},
  title = {{Modeling Gas Dynamic Effects in Shock-Tubes for Reaction Kinetics Measurements}},
  booktitle = {53rd},
  year = {2015},
  doi = {10.2514/6.2015-0414},
}
Figure