Fundamental Structure of High-

Speed Reacting Flows: Detonation

Jason Burr, Robert Fievisohn,

and
Ken H. Yu
Department of Aerospace Engineering
University of Maryland
College Park, MD

2014 AFOSR-ARO Basic Combustion Research Reviews

Arlington, VA; June 2-5, 2014
 Rotating Detonation Engine (RDE)
U N I V E R S I T Y O F M A R Y L A N D

Detonation Engines: Pressure-Gain Combustor

or Constant Volume Cycle
Simplicity in design
High thermodynamic efficiency
RDE: Detonation wave(s) spinning in an
annular tube fed from one end
Higher frequency than PDE
Wave propagation normal to the flow AFRL Experiment
direction (Naples et al. AIAA 2013-0278)
NRL Simulation
Active R&D (Schwer et al. AIAA 2014-1014)
with engine 7.6mm detonation channel
configuration
3.18mm
AFRL,
slot jet 80, 2.54mm
NRL,
holes
NPS, etc.
Air Hydrogen Air
 Rotating Detonation Wave Structure
(NRL Simulation; Kailasanath & Schwer, 2012)
U N I V E R S I T Y O F M A R Y L A N D

A- Detonation Wave
B- Oblique Shock Wave
C- Material Slip Line between freshly detonated
products and older products
D- Secondary Shock Wave
E- Mixing Region between fresh detonable mixture
and detonated gases
F- Region with blocked micro-nozzles,
G- detonable mixture injected from micro-nozzles

Kailasanath & Schwer, 2012 Int. Workshop Schwer & Kailasanath AIAA 2012-0617
on Detonation for Propulsion Feedback into Mixture Plenums in RDE
 Basic Nature of PDE and RDE
U N I V E R S I T Y O F M A R Y L A N D

Key Features of Detonation Engines PDE

Confined by rigid walls from all sides
transient and inherently unsteady
;

for steady state

PDE
Frequency ~ 101-102 Hz
Ignition every cycle, followed by DDT RDE
Detonation wave propagates in the Confined by two side walls,
direction of the flow (in-stream) permeable floor, and open ceiling
;
RDE
for steady state
Frequency ~ 103-104 Hz
Ignition once, only at starting
Detonation wave propagates normal to
the direction of the flow (cross-stream)
 Unresolved Research Issues in RDE (1)
U N I V E R S I T Y O F M A R Y L A N D

Fundamental structure of the governing flow field

Good progress in RDE development and CFD research
Critical gap in high-quality experimental data from basic research
Incomplete understanding of the limiting physics that may affect the operation
Critical height of detonable mixture layer (note: cell size dependent on local
mixture ratio)
Limiting flow rates for operation (note: not the design parameter dependence,
but rather a limiting factor, such as minimum energy rate for sustenance)
Incomplete understanding of mode hopping phenomena
Maximum delay time or minimum frequency (will this be controlled by auto-
ignition and DDT, or something else such as acoustic energy or entropy ?)
Stability criterion at given flow rate (will this be driven by device variation or
other geometric relation, such as Ljet/Lperim ~ Vinj/Vdet?)
 Unresolved Research Issues in RDE (2)
Contd
U N I V E R S I T Y O F M A R Y L A N D

Turbulent mixing of jets (fuel, oxidizer, products)

Premixed injector .vs. Separate Injection
Effects of jet height on wave speed
How much does the wave speed (average tangential component) get affected by
the degree of mixing between reactants/products or fuel/air/products?
Effects of jet velocity and residual products (i.e., jet-to-wall area ratio)
Azimuthal motion, secondary circulations, etc.
Effects of vorticity and localized swirl
There is strong vorticity production at the wave edge; Also, swirl may be present
due to the introduction of tangential velocity component by pressure gradient
(counter the rotation) and detonation wave (toward the rotation)
Effect of curvature
In average, the wave front near the outer wall needs to propagate faster than the
wave front near the inner wall. Will this be influenced by the wave tilt angle (r) or
the detonation wave speed Vdet(r)?
 Research Objectives
U N I V E R S I T Y O F M A R Y L A N D

Experimentally investigate the nature of detonation wave

propagating across the transversely injected reactant flows
Determine fundamental detonation wave and flow structure in RDE

Enhance basic understanding of rotating detonation engine (RDE)

operation by conducting detonation experiments in a linear
channel (analogous to a single-cycle unwrapped RDE)
Shed more lights on critical jet height, flow rates, injector-wall area
ratio, etc.
Build experimental database in a linear-channel transverse-pulse
detonation engine for future CFD validation
Investigate details of reactant injection and mixing at various initial
conditions and boundary conditions
 Design of Linear Model Detonation Engine Injector
U N I V E R S I T Y O F M A R Y L A N D

AFRL RDE baseline

injection (Schauer,
Fievisohn, etc.)

NRL RDE premixed

micro injection system
(Schwer & Kailasanath)

Linear Model Detonation Engine

injectors (for this study)
Recessed premixing tubes
Easily reconfigurable
Good visual access
Suited for systematic study
1~2 cycles per run
 Linear Model Detonation Engine Setup
U N I V E R S I T Y O F M A R Y L A N D

Simplified Configuration that Relevant Injector Dimensions

preserves the essential (AFRL Baseline: 2.54-mm fuel injector,
physics of the system 7.62-mm channel width, 5.95-mm spacing;
Detonation wave propagating NRL Baseline: premixed simulation)
through crossflow of reactants
Key Features of the setup oxidizer
Detonation wave initiated fuel
with single-shot PDE
Controlled fueling options
Fast-acting solenoid valves
Shock-tube driven injection
Modeling operational
characteristics
First cycle detonable gas
injection into air/inert gases
Subsequent cycle injection
into previous cycle products
and a wave from second
 Design of Experimentation and Setup
U N I V E R S I T Y O F M A R Y L A N D

Repeatable physical events

Timing-controlled data
acquisition

PDE

0.10 Dia Injectors

1.00 Premixing length
4 injectors/ inch
 Schlieren images of helium jet injection
U N I V E R S I T Y O F M A R Y L A N D
 Detonation Wave Initiation with PDE
U N I V E R S I T Y O F M A R Y L A N D

In the absence of detonable mixture, a

detonation wave turns into a shock wave
upon exiting the PDE tube
Measure propagation distance at various
angles as a function of t, and compare

Image taken at time t After Ignition

Reactants: stoichiometric H2-O2

r h mixture
Procedure: pre-fill for sec at ~0.1
g/sec, stop the reactants flow, and
apply spark ignition
x Reference Timing: w.r.t. ignition event
 Shock Wave radiating away from PDE exit
U N I V E R S I T Y O F M A R Y L A N D
 Correlation with Shock Travel Distance
U N I V E R S I T Y O F M A R Y L A N D
 PDE Wave discharging over LMDE
U N I V E R S I T Y O F
(no injection)
M A R Y L A N D

0 0+20s 0+38s 0+56s

0+72s 0+91s 0+127s 0+149s

 LMDE Injection Hydrogen only
U N I V E R S I T Y O F M A R Y L A N D
 LMDE Injection Hydrogen only
U N I V E R S I T Y O F M A R Y L A N D
 LMDE Injection Hydrogen only
U N I V E R S I T Y O F M A R Y L A N D
 LMDE Injection 50/50 Mixture of O2-He
U N I V E R S I T Y O F M A R Y L A N D
 LMDE Injection Control
U N I V E R S I T Y O F M A R Y L A N D

Fuel: H2 Oxidizer: O2-He

 PDE discharge into LMDE (H2-O2/He)H ~ 12 mm
U N I V E R S I T Y O F M A R Y L A N D

H2-O2/He injection
LMDE Injection
Reaction
2H2 + O2 + He 2H2O + He
Cell Width: ~ 1.2 mm H2-O2/He injection
Injector
Injection Height: H ~ 12 mm
when detonation wave arrives
H/ ~ 10, H/D ~ 5
Re ~ 3105 steady-state design,
actual would be much less
Results
Inconclusive; possibly near the He injection only
limit
Steepening of the wave front in x-
direction suggests amplification
But no noticeable difference with
the inert He injection case
 PDE discharge into LMDE (H2-O2) H ~ 12 mm
U N I V E R S I T Y O F M A R Y L A N D

H2 injection
LMDE Injection
Reaction
2H2 + O2 2H2O
Cell Width: ~ 1.4 mm H2-O2 injection
Injector
Injection Height: H ~ 12 mm
when detonation wave arrives
H/ ~ 8.6, H/D ~ 5
Re ~ 3105 steady-state design,
actual would be much less
He injection only
Results
Failed to maintain the detonation
wave front
No noticeable difference with the
inert He injection case
 PDE discharge into LMDE (H2-O2) H ~ 25 mm
U N I V E R S I T Y O F M A R Y L A N D

H2 injection
LMDE Injection
Reaction
2H2 + O2 2H2O
Cell Width: ~ 1.4 mm
H2-O2 injection
Injector
Injection Height: H ~ 25 mm when
detonation wave arrives
H/ ~ 18, H/D ~ 10
Re ~ 3105 steady-state design,
actual would be much less
Results H2-O2 injection
Detonation wave front appears to be
sustained in the forward direction
Wave front speed (x-direction) is
faster than in the inert gas injection
case
 PDE discharge into LMDE (H2-O2) H ~ 25 mm
U N I V E R S I T Y O F M A R Y L A N D

Injection height: H=25 mm; D=2.5mm, Cell size: ~ 1.4mm (H/ ~ 18, H/D ~ 10)
 PDE discharge into LMDE (H2-O2) H ~ 25 mm
U N I V E R S I T Y O F M A R Y L A N D

LMDE Injection
Reaction
2H2 + O2 2H2O
Cell Width: ~ 1.4 mm
Injector
Injection Height: H ~ 25 mm when
detonation wave arrives
H/ ~ 18, H/D ~ 10
Re ~ 3105 steady-state design,
actual would be much less
Results
Detonation wave front appears to
be sustained in the forward
direction
Wave front speed (x-direction) is
faster than in the inert gas injection
case
 PDE discharge into LMDE (H2-O2) H ~ 40 mm
U N I V E R S I T Y O F M A R Y L A N D

H2-O2 injection
H=40 mm; D=2.5mm (H/ ~ 29, H/D ~ 16)
H2-O2 injection w/o PDE

H2-O2 injection H2-O2 injection

 PDE discharge into LMDE (H2-O2) H ~ 40 mm
U N I V E R S I T Y O F M A R Y L A N D

LMDE Injection
Reaction
2H2 + O2 2H2O
Cell Width: ~ 1.4 mm
Injector
Injection Height: H ~ 40 mm when
detonation wave arrives
H/ ~ 29, H/D ~ 16
Re ~ 3105 steady-state design,
actual would be much less
Results
Revealed irregularly shaped wave
front - evidence of blast wave
Significant amount of amplification
suggests detonation transition would
be successful with side confinement
 Summary of Year 1 Progress
U N I V E R S I T Y O F M A R Y L A N D

New experimental framework for studying detonation wave propagating in a

narrow channel into a crossflow of reactants has been constructed and
experiments are underway. The objectives are to study the fundamental structure
in RDE flowfield and enhance our understanding of the physics. Two new
students, including one with a fellowship, are assigned to the investigation,
covering the experimental and analytical studies.
Design, fabrication, and preliminary testing of Linear Model Detonation Engine
(LMDE) is completed.
Development and testing of PDE for driving the LMDE rig is completed.
Timing sequence for synchronizing PDE and LMDE has been tested without any
confinement. The present results are being processed for preparing the future
tests with windows in place.
Preliminary tests indicate the present approach may be well suited for studying
physical mechanisms for RDE relevant issues such as critical height of reactant
jets, mode-hopping phenomenon, injection velocity and level of premixing.
 Additional Conclusions
U N I V E R S I T Y O F M A R Y L A N D

Under certain conditions in the present tests, even without any flow
confinement, it was clearly shown that shock wave could be effectively
accelerated in the forward direction by the reaction from the transverse jets. It
appears to be dependent on the reactant jet height, among other factors.
In the present configuration of unconfined reactant jets, a jet height on the
order of 10 or greater was needed to continue driving the shock front in the
initial wave direction. This result is likely to be configuration specific since it
would be affected by the reactant jet blockage ratio and the rate of ambient
fluid entrainment as well as the amount of fuel present.
In the present system, timing of detonation wave arrival at the PDE exit is not
consistent enough to be used as a controlled input to LMDE, most likely due to
DDT phenomenon inside the PDE tube not being accurately predictable. This
uncertainty (order of ~100s) could be reduced by introducing a transition
enhancement device or using direct initiation.
 Plans for Year 2
U N I V E R S I T Y O F M A R Y L A N D

Investigation of the fundamental structure of RDE flow-field

Confined LMDE experiments (unwrapped RDE channel) with side walls in
place for measuring basic flow structure
Pressure diagnostics on side walls, High-speed visualization,
chemiluminescence (OH*)
Study of specific and relevant phenomena
Critical jet height in LMDE experiment
Parametric investigation of flow rate or velocity dependence
Facility upgrade
Higher flow rate handling capability
PDE improvement for minimizing the timing uncertainty
Additional PDE for successive cycle driving
2-D analysis of RDE
Simplified flowfield with periodic boundary and inert gas upper layer
Method of Characteristics based approach for shock-slip line interaction
 Acknowledgment
U N I V E R S I T Y O F M A R Y L A N D

Dr. Chiping Li, AFOSR

Dr. K. Kailasanath, NRL

Dr. Fred Schauer, AFRL
 Fundamental Structure of High Speed Reacting Flows: Transverse
Detonation Wave
PI: Ken Yu --- Grant No.: FA9550-13-1-0080 --- Univ. of Maryland
Transverse Propagation of Detonation Wave in a Linear, Optically-Accessible, Rectangular Waveguide

TECHNICAL APPROACH: PROGRESS:

INTRODUCTION

CURRENT PROGRESS
GOAL Linear transverse detonation wave
Support Rotating Detonation Engine Linear Model Detonation Engine (LMDE)
channel with full optical access:
(RDE) development by providing better Canonical configuration with simplified
- preliminary design and analysis
understanding of the fundamental wave geometry but essential physics
completed
structure and building database for
- parts acquisition, machining &
CFD validation
fabrication in progress
- PDE initiator built & tested
BACKGROUND
Renewed interest in RDE concept due
to potential performance gain
Incomplete understanding of the actual
wave structure involving complex flow Experimentally simulate one or more cycles of RDE
features operation in an unwrapped linear channel for high-
Lack of high-quality visualization data quality flow viz. and detailed diagnostics air
fuel
due to complex optical setup involving - pulsed detonation tube as wave initiator
annular passage way LMDE v.1
- reflected wave to simulate the continuing mode
Need for experimental data that can be Study fundamental details of detonation structure PDE initiator
used for new physical insights as well propagating transversely across the flow direction in a
as for CFD validation purpose confined passage way

EXPECTED RESULTS
- high-speed schlieren visualization
OBJECTIVES
OBJECTIVES

- synchronized chemiluminescence measurement

Investigate fundamental structure of Planned Impact
transverse detonation structure Expected Milestones in Year 1: Obtain high-quality experimental data
representing of RDE operation Fabrication of Linear Model Detonation Engine on transverse propagation of
Characterize key flow features, Transient flow path mapping and mixing characterization detonation wave in cross flow
understand their behavior, and provide Development of high-speed flow visualization system Build scientific database for CFD
new insights customized for LMDE validation supporting RDE
development