JID:AESCTE AID:4689 /FLA [m5G; v1.243; Prn:26/09/2018; 16:49] P.

1 (1-8)
Aerospace Science and Technology ••• (••••) •••–•••

1 67
Contents lists available at ScienceDirect
2 68
3 69
4 Aerospace Science and Technology 70
5 71
6 72
7
www.elsevier.com/locate/aescte 73
8 74
9 75
10 76
11
12
Characterization of induction and transition methods of oblique 77
78

detonation waves over dual-angle wedge ✩

OF
13 79
14 80
15 b a c 81
Bikalpa Bomjan , Sudip Bhattrai , Hao Tang
16 82
a 83
17 Department of Mechanical Engineering, Institute of Engineering – Pulchowk, Tribhuvan University, Chakupat, Pulchowk-44700, Nepal
b
18 Flight Operations Engineering and Planning Division, Buddha Air, Nepal 84

RO
c
Jiangsu Province Key Laboratory of Aerospace Power System, College of Energy and Power Engineering, Nanjing University of Aeronautics and Astronautics, 85
19
29-Yudao St., Nanjing-210016, China
20 86
21 87
22 88
a r t i c l e i n f o a b s t r a c t
23 89
24 Article history: 90
The oblique detonation wave (ODW) induction and initiation lengths can be reduced significantly by the

DP
25 Received 10 December 2017 use of a dual-angle wedge with two subsequent, high and low, angles. In this study, the dual-angle wedge 91
26 Received in revised form 6 May 2018 is deflected to different angles at different positions along the wedge and the resulting effects on the 92
Accepted 23 July 2018
27 shock-to-detonation transition are observed. One-dimensional analytical modeling of flow conservation 93
Available online xxxx
28 equations with finite-rate multi-step reaction is carried-out to obtain reactive flow properties over plain 94
29 Keywords: and dual-angle wedges. Numerical study was carried out in OpenFOAM to observe the induction and 95
30 Shock-induced shock-to-detonation transition characteristics. Two different transition mechanisms were observed over 96
31 Combustion dual-angle wedges. The first is an abrupt-type transition from an oblique shock wave (OSW). While the 97
TE
Hypersonic second mechanism involves a brief induction process followed by an intermediate span of shock-induced
32 98
Expansion waves combustion, which leads to a smooth transition into an ODW. The ODW is formed over the second
33 OpenFOAM 99
wedge-angle in both cases. The ODW induction and initiation lengths can be reduced significantly by
34 100
initially inclining the wedge at a higher angle over only a very short span of the wedge. The second
35 101
wedge-angle primarily affects the ODW initiation length and downstream flow properties. The position
36 of wedge deflection was identified as a critical geometric parameter which can affect the ODW induction, 102
EC

37 transition and formation properties. 103

38 © 2018 Elsevier Masson SAS. All rights reserved. 104
39 105
40 106
41 107
RR

42 108
43 109
44 1. Introduction a sufficiently high post-OSW temperature the induction process is 110
45 kinetically-controlled, where the chemical kinetics primarily drives 111
46 An attached oblique shock wave (OSW) can be used to ig- the change in flow properties in the induction region. While if the 112
47 nite a supersonic/hypersonic fuel-air mixture flow over a ramp post-OSW temperature is too low so that the elevated pressure 113
CO

48 or a wedge, leading to the formation of an oblique detonation behind it governs the chemical kinetics, the induction process is 114
49 wave (ODW). Under suitable conditions, the ODW is formed af- wave-controlled. For a kinetically-controlled induction process, the 115
50 ter a brief induction process and it is attached to the induction shock-to-detonation transition is smooth, which is characterized by 116
51 OSW. The reaction induction length, or induction delay, is the flow a curved induction OSW [6,7]. Whereas, for a wave-controlled in- 117
52 distance between the temperature increase behind the OSW and duction process, the shock-to-detonation transition is characterized 118
53 the beginning of the rapid chemical reaction, which depends on by an abrupt transition at a multi-wave point [6]. Hence the type 119
UN

54 the temperature and pressure behind the OSW [1–4]. The ODW of induction process that occurs behind an OSW can also deter- 120
55 initiates at a triple-point where the induction process fully cou- mine the shock-to-detonation transition mechanism [5,6,8–14]. 121
56 ples with the induction OSW. Recent studies [5,6] have shown that The shock-to-detonation transition structure is commonly ex- 122
57 the induction process can either be kinetically-controlled, or wave- pressed in the simplified form of the traditional planar ZND det- 123
58 controlled. In the presence of a high inflow Mach number and/or onation model. However, such planar structures have been shown 124
59 to only exist closer to the ODW initiation region, and the ODW 125
60 structure becomes unstable further along the leading shock. This 126
61 ✩
This work was supported by the National Natural Science Foundation of China
cellular instability has been discussed extensively in recent stud- 127
62 (NSFC No. 51576098). ies on the structure and stability of ODWs [11,12,15,16], and it is 128
63 E-mail address: [email protected] (S. Bhattrai). also found to play an important role on the ODW transition and 129
64 130
https://doi.org/10.1016/j.ast.2018.07.038
65 131
1270-9638/© 2018 Elsevier Masson SAS. All rights reserved.
66 132
 JID:AESCTE AID:4689 /FLA [m5G; v1.243; Prn:26/09/2018; 16:49] P.2 (1-8)
2 B. Bomjan et al. / Aerospace Science and Technology ••• (••••) •••–•••

1 stability. Such transient instabilities were studied in [17], using the 67

2 Method of Characteristics and unsteady simulations [18]. It was 68
3 found that upon the formation of an overdriven ODW, the shock 69
4 and reaction complex exhibit a transient spatial oscillation that 70
5 leads to the generation of transverse waves. The resulting cellular 71
6 structure is comprised of the oscillating complex of induction and 72
7 transverse shock waves, triple points and ODWs [19]. In addition to 73
8 understanding of the planar ZND model of an ODW, the ODW cel- 74
9 lular structures provide a better understanding of real ODW flows, 75
10 so such inherent phenomena are necessary to be accounted for in 76
11 numerical studies [15]. 77
12 The ODW literature from the past few decades provide a large 78

OF
13 volume of studies on the limits of hydrodynamic stability of ODWs 79
14 [1,2,10,14], as well as a growing number of numerical studies re- 80
15 lated to the structure and stability of ODWs for propulsive appli- 81
16 cations [20,21]. The formation of an ODW in a hypersonic com- 82
17 bustion system can lead to significant advantages in terms of the 83
18 combustion length required, and there are possibilities for appli- 84

RO
19 cation of the detonation phenomena in hypersonic air-breathing 85
20 vehicles such as the oblique detonation wave engine (ODWE) and 86
21 shock-induced combustion ramjet (shcramjet). 87
22 For a given upstream flow condition, the hydrodynamic stabil- 88
Fig. 1. Schematic representations of (a) the formation of an ODW over a dual-angle
23 ity of an OSW requires that the wedge angle is below the limiting wedge, and (b) the position of wedge deflection at 2 mm (P.1) and 4 mm (P.2).
89
24 angle, beyond which the OSW detaches and propagates upstream 90

DP
25 in the flow. The maximum allowable wedge angle for an ODW is, 91
26 however, lower than the OSW detachment angle because of heat and the resulting ODW angle. The reason for this difference was 92
27 release behind the ODW. Analytical studies of finite-rate reaction explained to be the ideal flow conditions and instantaneous heat 93
28 chemistry and heat release can provide a reliable prediction of release assumption in the theoretical analysis. 94
29 shock-induced combustion and flow properties, to guarantee the In a practical ODW combustor, achieving proper premixing of 95
30 hydrodynamic stability of ODWs in practical ODW combustors [1, the fuel-air mixture and/or effective ODW combustion within a 96
31 2,22–24]. given combustor geometry can pose significant challenges. Studies 97
TE
32 It is of interest for practical applications to understand the related to the effects of the flow parameters, mixture properties 98
33 mechanisms of formation and characteristics of ODWs, as well [31], or inhomogeneities in flow conditions [32] have received the 99
34 as, to determine the suitable geometric and operational condi- much-needed attention recently. Similarly, the geometrical meth- 100
35 tions [25–29]. Furthermore, it is also desirable to explore methods ods [25,30] for obtaining ODWs with desirable formation and com- 101
36 that can allow for the formation of ODWs at low wedge-angles as bustion characteristics is an interesting problem in ODW research. 102
EC

37 well as with reduced ignition delay [30]. In particular, Chapman– This study intends to characterize the formation mechanisms of 103
38 Jouguet (CJ) or near-CJ ODWs provide desirable combustion prop- ODWs over dual-angle wedges, and study the induction and initi- 104
39 erties at low values of ODW angle, and studies related to the for- ation properties, with respect to the change in wedge geometry. 105
40 mation of CJ or near-CJ ODWs have been previously carried out to The dual-angle wedge geometry is varied with respect to: 106
41 understand their physics and stability [14,17,20,25], and to trans- (1) deflection position of the wedge, and 107
(2) the angles of the wedge.
RR

42 late these understanding to practical applications. 108

43 Pimentel et al. [20] described a method to obtain a CJ ODW by 109
44 initially triggering an overdriven ODW by a wedge and suddenly 2. Analytical study of dual-angle wedge 110
45 lowering it to an angle parallel to the ODW upstream flow direc- 111
46 tion, such that, the angle of an initially overdriven ODW relaxes A dual-angle wedge is a conceptually simple combustor geom- 112
47 down to the value corresponding to the CJ ODW. Similarly, Verrault etry where a wedge is inclined at high and low overdriven angles, 113
CO

48 et al. [17] used a dual-stage wedge to modify the overdrive of an respectively. The first wedge-angle is relatively short in span, over 114
49 ODW. The expansion fan, which is generated because of the abrupt which the chemical induction takes place behind a stronger OSW. 115
50 deflection of the wedge surface, interacts with the overdriven ODW The wedge is then deflected to the second wedge-angle, over which 116
51 and decreases the pressure and temperature downstream of the the shock-to-detonation transition takes place and a standing ODW 117
52 ODW. This leads to an increase in Mach number in the burned gas is formed. This concept is shown schematically in Fig. 1. The po- 118
53 region, and also to a decrease of the wedge-ODW angle. sitions of wedge deflection are designated as P.1 and P.2 in Fig. 1 119
UN

54 Another recent study by Bhattrai et al. demonstrated a method, that are at 2 mm and 4 mm, respectively, horizontal from the cor- 120
55 using a dual-angle wedge, to reduce the ODW initiation length by ner of the wedge. The second wedge-angle is greater than the CJ 121
56 inclining the wedge initially at a higher angle over only a very wedge-angle, determined using analytical calculation for the post- 122
57 short span of the wedge [30], without the need for a large wedge OSW flow properties. 123
58 length as required in Ref. [20]. A non-uniform wedge geometry In the analytical study, the combustion is assumed to take place 124
59 that allows for varying degrees of overdrive of the initial OSW to in a one-dimensional steady flow with finite-rate chemistry at time 125
60 achieve early induction, but with the formation of ODW at low steps of 2e-10 s. Also, the combustion process is assumed to take 126
61 turning angles, was utilized to obtained desirable ODW proper- place in two steps – first the chemical process takes place, which is 127
62 ties for a practical ODW combustor. A method to obtain a near-CJ then followed by the thermophysical processes at each time step. 128
63 ODW that can have a minimum value of βw and a reduced total The finite-rate chemistry is modeled in MATLAB using a 9 species, 129
64 pressure loss, for any given amount of heat release, was discussed. 12-step reaction mechanism from Ref. [33]. The reduced chemical 130
65 However, the study found significant differences between numer- reaction mechanism consists of six irreversible and six reversible 131
66 ical and theoretical predictions for heat release behind the ODW reactions, simplified from the 21 step “San Diego mechanism” [34] 132
 JID:AESCTE AID:4689 /FLA [m5G; v1.243; Prn:26/09/2018; 16:49] P.3 (1-8)
B. Bomjan et al. / Aerospace Science and Technology ••• (••••) •••–••• 3

1 67
2 68
3 69
4 70
5 71
6 72
7 73
8 74
9 75
10 76
11 77
12 78

OF
13 79
14 80
15 81
16 82
17 83
18 84

RO
19 85
20 86
21 87
22 Fig. 2. Evolution of temperature obtained from the analytical modeling of finite-rate 88
23 reaction and RH equations for planar wedges. 89
24 90

DP
25 and validated for detonations under the conditions of practical in- 91
26 terest with reasonable accuracy [33]. 92
27 The finite-rate reaction progression results in a change in over- 93
28 all mixture density and pressure, computed from the change in 94
29 species concentration at each time step. The effect of the result- 95
30 ing heat release on flow parameters are obtained by solving the 96
31 basic conservation equations, using Rankine–Hugoniot (RH) anal- 97
TE
32 ysis method, for non-reacting one-dimensional steady flow. The 98
33 method assumes that the heat release takes place prior to combus- 99
34 tion (pressure) wave, and the resulting change in flow parameters 100
35 behind the combustion wave can be obtained from the solution of 101
36 conservation equations. 102
EC

37 Upon equilibrium, the induction and reaction completion times 103

38 can be obtained from the reaction progression plots, for. e.g., as 104
39 shown in Fig. 2. The induction time, τign , is defined as the time 105
40 taken for the mixture temperature to rise from initial tempera- 106
41 ture to 1% of the equilibrium temperature, and the reaction time 107
RR

42 τrxn is defined as time taken for the mixture temperature to rise Fig. 3. Analytically obtained combustion length scales for dual-angle wedges, where
108
43 from 1% to 99% of the equilibrium temperature [24]. These time the expansion processes are assumed to take place suddenly at 2 mm (P.1) and
109
44 periods are important parameters in the current study, since they 4 mm (P.2) for (a) 21◦ –12◦ and (b) 21◦ –15◦ . 110
45 can generally provide the length scale estimation for ODW initi- 111
46 ation. The prediction accuracy of the current analytical model is However, the equilibrium temperatures are found to depend only 112
47 determined by comparing the results provided in Ref. [33]. For on the second wedge-angle. As a result, the reaction completion 113
CO

48 the modeling of finite-rate chemistry at constant pressure, using time for a the 15◦ second-wedge angle is lower-irrespective of the 114
49 the post-OSW temperature and pressure, the results obtained for deflection position. In addition, it can be noted that the effect of 115
50 temperature and species concentration variations agree accurately the wedge deflection on induction lengths for 21◦ –12◦ dual-angle 116
51 with the results in Ref. [33]. Since the post-OSW induction pro- wedge at P.2 is comparable to the 21◦ –15◦ dual-angle wedge at 117
52 cess occurs essentially at a constant-pressure, the induction times P.1, which is close to 10 mm for both the cases. This shows the 118
53 obtained from the chemistry solver provide reliable estimates for similarity in effects of the position of deflection and the second 119
UN

54 ODW induction delay. Furthermore, the inclusion of RH equations wedge-angle on induction length, although their matching values 120
55 into the chemistry solver provides the estimates for the reaction are merely coincidental. 121
56 completion periods, accounting for jumps in pressure and density Results in Table 1 show that, for a given wedge of 80 mm span, 122
57 across the combustion waves, as shown in Fig. 2. From the fig- the non-dimensional heat release for 15◦ plain wedge-angle is the 123
58 ure, it is found that the induction length decreases from 58 mm to highest. Also, heat release for the 21◦ –15◦ dual-angle wedge is 124
59 5 mm with an increase in wedge angle from 12◦ to 21◦ . higher than that for the 21◦ plain wedge, which implies a better 125
60 The effects of changing the second wedge-angle and position of combustion efficiency for the earlier case. Based on the results in 126
61 deflection of a dual-angle wedge is shown in Fig. 3. Keeping the Fig. 2, Fig. 3 and Table 1, it can be seen that the first wedge-angle 127
62 first and second wedge-angles constant at 21◦ –15◦ or 21◦ –12◦ , a of 21◦ decreases the induction length—significantly in comparison 128
63 shift in the deflection position from P.1 to P.2 reduces the induc- to 15◦ and 12◦ . On the other hand, decreasing the wedge-angle 129
64 tion length. This is because the flow residence time is longer over to lower inclinations of 15◦ and 12◦ can in fact lead to a larger, 130
65 the first segment of the wedge where the temperature and pres- or comparable, amount of heat release than at 21◦ . Therefore, in 131
66 sure behind the primary OSW are higher than the deflected OSW. terms of both the reduction in induction and initiation lengths, 132
 JID:AESCTE AID:4689 /FLA [m5G; v1.243; Prn:26/09/2018; 16:49] P.4 (1-8)
4 B. Bomjan et al. / Aerospace Science and Technology ••• (••••) •••–•••

1 Table 1 67
2 Comparison of combustion and ODW properties for plain and dual-angle wedge of 8 cm hori- 68
3 zontal span at inflow conditions of Mach 6, 800 K and 50 kPa. 69
4 θ (deg) Non-dimensional finite-rate heat release (q = Q /C p T ) βODW (deg) 70
5 (1) 12 1.574 31.2 71
6 (2) 15 1.981 35.5 72
7 (3) 21 1.729 38.0 73
(4) 21–12 at P.1 1.867 33.4
8 74
(5) 21–12 at P.2 1.869 33.4
9 (6) 21–15 at P.1 1.825 34.0 75
10 (7) 21–15 at P.2 1.825 34.0 76
11 77

∂(ρ U )
12 78
and the combustion efficiency, the analytical results show that the  ) = −∇ p
Momentum : + ∇(φ U (2)

OF
13 79
14
dual-angle wedge presents a significant advantage over a plain ∂t 80
15
wedge. ∂(ρ E ) 81
16
There are two main observations made from the analytical Energ y : + ∇(φ H ) = 0 (3) 82
∂t
17 study: 83
∂(ρ Y i )
18 (1) For fixed first and second wedge-angles, the position of Species : + ∇(φ Y i ) = ẋi W i (4) 84

RO
19 wedge deflection has the effect of changing the induction and reac- ∂t 85
20 tion completion lengths, which decrease as the deflection position  1 86
1 2 2 p
21 moves downstream. The equilibrium temperature, along with the Where, E = e + 2
U = Y i e i + U is total energy, H = E + ρ , 87
2
22 ODW angle and non-dimensional heat release, is primarily depen- i 88
dent on the second wedge-angle. is the total enthalpy used in AUSM+ energy flux calculation [40]
23 89
(2) Despite the short length, the value of the first wedge-angle  is the convective face flux. The above set of equations
and φ = ρ U
24 90
is solved in combination with the equation of state for perfect gas

DP
and the flow properties behind the primary OSW, strongly affects
25  Yi 91
26 the ODW induction length. Apart from this, observation of the mixture, p = ρ R T , to completely define the inviscid reac- 92
27 parameters in Fig. 2, Fig. 3 and Table 1 show that the first wedge- Wi 93
i
28 angle does not substantially determine the final ODW properties. tive flow field. In the species conservation equation, Y i is mass 94
29 In addition, Table 1 shows that the cases with lower second- fraction of the chemical species i, and ẋi W i is the reaction source 95
30 wedge angles have higher heat release values at lower ODW an- term where W i is the molecular weight of the species i and ẋi is 96
31 gles, which are desirable properties for higher combustion effi- the specie molar production rate. The reaction rate-constants k for 97
TE
32 ciency and pressure recovery. The analytical findings also indicate elementary reactions are calculated using the Arrhenius law, 98
33 that a near-CJ ODW can be formed over a dual-angle wedge when 99
the overdriven second-wedge angle approaches the CJ wedge an-
34 k = AT β exp(− E / R T ) (5) 100
35 gle. 101
36 The kinematic mechanism for the H2 -air mixture considers a 102
EC

37 3. Numerical approach 9-species [O2 , H, OH, O, H2 , H2 O, N2 , HO2 , H2 O2 ] reaction mecha- 103

38 nism with 12 elementary reactions, from Ref. [33], as used in the 104
39 For numerical modeling, the Euler equations for 2-D unsteady analytical modeling. The reaction rate k and molar production rate 105
40 compressible reactive flow are solved with a solver written in xi are calculated using the CHEMKIN package. The thermodynamic 106
41 OpenFOAM. In an ODW flow, the effects of viscosity and thermal data is supplied with the CHEMKIN file as piece-wise polynomial 107
RR

42 dissipation are confined to the thin boundary layer and the change functions of temperature [41]. 108
43 in boundary layer thickness is small [10,35,36]. The effect of flow The face convective flux values are calculated by AUSM+ flux 109
44 viscosity on ODW characteristics was studied in Ref. [35,37], and discretization method. The cell-to-face extrapolation of primitive 110
45 the boundary layer was shown to have little effect on the main variable is obtained by extrapolating the cell centered variables 111
46
wave structure [12]. Similar numerical studies on the formation to face centers using a second-order Taylor series expansion. A 112
47
and structure of detonations [38], as well as, on the formation, cell limited gradient scheme is used for scalar gradient calcula- 113
CO

48
stability, and structure of ODWs have neglected the effects of vis- tion, while a limited-linear TVD scheme is used for interpolation 114
49
cosity, mass diffusion and thermal conduction [6,11,15,17,19,31]. of primitive variables based on the direction of flux at each face. 115
50
Also, since the wedge is deflected away from the flow, the for- Numerical study was carried out in domains with wedges of 50 116
51
mation of a separation shock observed in a double-wedge flow is mm, 60 mm, 80 mm and 120 mm horizontal spans, as shown in 117
52
avoided [39]. Therefore, the possibility a boundary layer interac- the schematic in Fig. 1, with the wedge starting at the origin. The 118
53
tions near the deflection region of the dual-angle wedge is ignored dual-angle obstacle surface of the wedge is modeled as an inviscid 119
UN

54
for computational simplicity, focusing on the inviscid ODW flow wall. A bottom symmetry surface is added upstream of the corner 120
55
characteristics. However, the presence of a separation region at the of wedge containing 10 cells in order to avoid numerical reflection 121
56
wedge corner can have a strong effect in the induction shock prop- from the inlet boundary. The inlet is modeled as a fixed-velocity 122
57
erties. Such complex flow problems are currently outside the scope inlet, and the outlet is modeled as an extrapolated boundary with 123
58
of our study, and merit a large volume of studies in ODW litera- zero-gradient conditions imposed on all scalar values at the outlet 124
59
ture in themselves in the future. Nevertheless, for all the discussion surface. The top surface is at freestream condition to avoid shock 125
60
within the scope of the current study, the authors believe that an reflection. The flow is a stoichiometric premixed H2 -air mixture 126
61
inviscid flow field can resolve the properties under investigation. having N2 mass fraction of 0.7435. 127
62 Mesh independence studies were carried-out at two wedge an- 128
The conservation form of the governing equations are formu-
63 gles of 18◦ and 21◦ . Five uniform mesh densities for a wedge of 129
lated as,
64 50 mm horizontal span inclined at 18◦ and stoichiometric H2 -air 130
65
∂ρ mixture inflow conditions of Mach 6, 800 K temperature and 131
66 Continuit y :  =0
+ ∇(φ) (1) 35 kPa pressure were used, with cell sizes, (x), of 0.25 mm, 132
∂t
 JID:AESCTE AID:4689 /FLA [m5G; v1.243; Prn:26/09/2018; 16:49] P.5 (1-8)
B. Bomjan et al. / Aerospace Science and Technology ••• (••••) •••–••• 5

1 Table 2 67
2 Comparison of initiation point and pressure change across an ODW at five uniform cell sizes for an 18◦ wedge at inflow conditions 68
3 of Mach 6, 800 K and 35 kPa. 69
4 x (mm) Location of ODW initiation point (mm) ODW peak pressure (kPa) ODW downstream pressure (kPa) 70
5 (1) 0.25 (28.5, 12.9) 384 323 71
6 (2) 0.125 (35.9, 16.6) 385 320 72
7 (3) 0.08 (38.0, 17.7) 410 345 73
(4) 0.06 (41.4, 19.2) 418 340
8 74
(5) 0.04 (43.4, 20.05) 424 334
9 75
10 76
11 77
12 78

OF
13 79
14 80
15 81
16 82
17 83
18 84

RO
19 85
20 86
21 87
22 88
23 89
24 90

DP
25 91
26 92
Fig. 5. Pressure results for 15◦ plain wedge at Mach 6, T = 800 K and p = 50 kPa.
27 (For interpretation of the colors in the figure(s), the reader is referred to the web 93
28 version of this article.) 94
29 95
30 the wedge is fully deflected to the second wedge-angle the OSW 96
31 slopes to a lower wave angle. As it is generally expected from this 97
TE
Fig. 4. Pressure variation across an ODW over a 21◦ plain wedge at inflow conditions
32
of Mach 6, 800 K and 35 kPa.
method, the YH2O and density flow field result in the figure shows 98
33 that the induction process progresses over the deflected ramp, and 99
34 the ODW is formed over the second wedge-angle—attached to the 100
0.12 mm, 0.08 mm, 0.06 mm, and 0.04 mm. The ODW peak pres-
35 deflected OSW. While the induction process initiates early on, near 101
sure and downstream pressure were measured at y = 23 mm and
36 the wedge deflection region, the reaction zone thickness gradu- 102
y = 18 mm, respectively. In addition, the ODW initiation length for
EC

37 ally grows until it intercepts the induction OSW, and a relatively 103
the five cases were observed, as given in Table 2.
38 abrupt transition into an ODW is observed. The average deflected 104
From the table, it can be seen that the peak pressure values
39
are close for x = 0.08 mm, 0.06 mm and 0.04 mm. The variation OSW angle in the numerical result is obtained to be 22.9◦ , which 105
40
of pressure across the ODW over 21◦ wedge angle for cell sizes is close to the theoretical OSW angle of 22.4◦ at 15◦ wedge angle. 106
41
x = 0.25 mm, 0.06 mm and 0.04 mm is shown in Fig. 4. The peak The deflected OSW remains effectively detached from the combus- 107
RR

42
pressure values for all the three cases are close to 420 kPa, while tion process prior to the transition. 108
43
the average ODW downstream pressure values for x = 0.06 mm As expected, even with the extremely short span of the first 109
44
and 0.04 mm lie close to 360 kPa towards the outlet. wedge-angle, the ODW initiation length is lower in comparison to 110
45
The cell sizes between 0.06 mm and 0.04 mm are typical in nu- the plain-wedge case in Fig. 5. Hence, as also observed analyti- 111
46
merical studies related to the physical characteristics of ODWs [5, cally, the high first wedge-angle only has an effect of expediting 112
47
31], as well as, to capture the fine cellular structures of ODWs [19]. the early induction. It is also important to note that despite the 113
CO

48
The fourth mesh density, with cell size x = 0.06 mm, is used for differences in induction region properties, the formation mecha- 114
49
all numerical simulations presented in the following sections. nism of the ODW in Fig. 6 appears characteristically identical to 115
50 the ODW formed over a plain wedge in Fig. 5. 116
51
4. Results and discussion In Fig. 6, the ODW induction length is around 10 mm, which 117
52 is close to the analytically obtained result. However, the induction 118
53 In the following section, the numerical results obtained from length values may deviate for different cases, and is slightly shorter 119
UN

54 OpenFOAM are presented, in light of the analytical results pre- for analytical results where the wedge deflection is assumed to 120
55 sented in Section 2. take place suddenly. Also, the reaction completion length for the 121
56 Fig. 5 shows the pressure flow field for a 15◦ plain wedge 21◦ –15◦ dual-angle wedge in Fig. 3(b) (deflection at P.1) provides 122
57 where the shock-to-detonation transition is observed to take place some indication for the ODW initiation length. However, the actual 123
58 abruptly, without the induction-OSW being affected by the down- ODW initiation length cannot be obtained from Fig. 3(b). 124
59 stream combustion processes. This is a simple case of an abrupt For a first wedge-angle of 21◦ , the pressure field results for a 125
60 shock-to-detonation transition, commonly observed in the over- second wedge-angle of 12◦ , deflected at P.1 and P.2, are shown in 126
61 driven ODW flows. Fig. 7. The case (a), with the wedge deflection at P.1, has an initi- 127
62 In Fig. 6, an ODW formed over a 21◦ –15◦ dual-angle wedge ation length significantly larger than that for case (b) that has the 128
63 deflected at P.1 is shown with pressure, density and H2 O mass wedge deflection at P.2. In the first case, which requires a wedge 129
64 fraction (YH2O ) flow fields. Similar to the illustration in Fig. 1, the of 120 mm horizontal span, the ODW initiation occurs close to 130
65 primary OSW angle gradually decreases due to the interaction with 100 mm. On the other hand, for the wedge deflected at P.2 the 131
66 expansion waves that form over the wedge deflection region. Once ODW initiation takes place over the second wedge-angle within a 132
 JID:AESCTE AID:4689 /FLA [m5G; v1.243; Prn:26/09/2018; 16:49] P.6 (1-8)
6 B. Bomjan et al. / Aerospace Science and Technology ••• (••••) •••–•••

1 67
2 68
3 69
4 70
5 71
6 72
7 73
8 74
9 75
10 76
11 77
12 78

OF
13 79
14 80
15 81
16 82
17 83
18 84

RO
19 85
20 86
21 87
22 88
23 89
24 90

DP
25 91
26 92
27 93
28 94
29 95
Fig. 7. Pressure flow field results for 21◦ –12◦ dual-angle wedge at Mach 6, T = 96
30
800 K and p = 50 kPa, showing an ODW initiation processes for wedge deflections
31 97
TE
at (a) P.1 and (b) P.2.
32 98
33 99
34 100
35 101
36 102
EC

37 103
38 104
39 105
40 106
41 107
RR

42 108
43 109
44 110
45 111
46 112
47 113
CO

48 114
Fig. 6. Pressure, density and YH2O results for ODW over a 21◦ –15◦ dual-angle wedge
49 115
at inflow conditions of Mach 6, 800 K and 50 kPa for wedge deflection at P.1.
50 116
51 117
52 50 mm span. This supports the interim conclusion from the analyt- 118
53 ical study of Section 2 that the induction and reaction completion 119
UN

54 lengths are highly sensitive to the deflection position. 120

55 However, the mechanism of ODW formation for the latter case 121
56 (P.2) in Fig. 7(b) is not the same as the case in Fig. 7(a), as well 122
57 as, the case in Fig. 6. One basic observation for the second ODW 123
58 case in Fig. 7 is that the shock-to-detonation transition is highly 124
59 smooth in comparison to the previous cases. The deflected OSW 125
60 gradually curves upwards as a result of the heat release behind it, 126
61 and a definitive point of transition into ODW is not as obvious as 127
62 in the earlier cases. 128
63 These phenomena can be more effectively clarified by a com- 129
64 parison of density field results in Fig. 8. Since, the flow density Fig. 8. ODW initiation mechanisms for 21◦ –12◦ dual-angle wedge at Mach 6, T =
130
65 decreases with the initiation of chemical reaction aft of the de- 800 K and p = 50 kPa, showing reaction progression in terms of the variation of 131
66 flected OSW, the density flow fields show the extent of detachment density for wedge deflections at (a) P.1 and (b) P.2. 132
 JID:AESCTE AID:4689 /FLA [m5G; v1.243; Prn:26/09/2018; 16:49] P.7 (1-8)
B. Bomjan et al. / Aerospace Science and Technology ••• (••••) •••–••• 7

1 In this study, several cases were observed for these phenomena 67

2 that appears when the dual-angle wedge is deflected from a point 68
3 where the reaction induction has started. A reference for when the 69
4 difference in transitions occur can be estimated based on the an- 70
5 alytical results in Fig. 3. As can be seen for both the 21◦ –12◦ and 71
6 21◦ –15◦ dual-angle wedges, the deflection position P.2 approxi- 72
7 mately coincides with the induction length of the 21◦ plain wedge 73
8 case. Hence, the smooth transition of the type in 21◦ –12◦ (at P.2) 74
9 case occur when the wedge deflection position lies close to the in- 75
10 duction length corresponding to the first wedge-angle. When the 76
11 deflection position is moved further downstream, the induction re- 77
12 gion shifts over to the first wedge-angle, until the ODW initiation 78

OF
13 takes place entirely over the first wedge-angle. 79
14 On the other hand, the typical abrupt transitions, as in Fig. 6, 80
15 Fig. 7(a) and Fig. 8(a), are obtained when the dual-angle wedge 81
16 is deflected relatively early in the induction process. The role of 82
17 higher first-wedge angle is purely to kick-start the induction pro- 83
18 cess early, and the characteristics of the ODW that form in the 84

RO
19 process are entirely dependent on the position of wedge deflec- 85
20 tion and the second wedge-angle. While the deflection position 86
21 was earlier found to strongly affect the induction length, it is 87
22 clear from the above discussion that it can also strongly affect the 88
23 shock-to-detonation transition mechanism and the ODW initiation 89
24 characteristics. 90

DP
25 These findings also have interesting practical implications. The 91
26 induction region can be accommodated within any reasonable ge- 92
27 ometrical length by adjusting the first wedge-angle, while at the 93
Fig. 9. OSW and ODW angles (figure to be replaced).
28 same time maintaining the desirable ODW properties over the sec- 94
29 ond wedge-angle. 95
30 96
between the deflected OSW and the ensuing combustion processes.
31 5. Conclusions 97
TE
32
In Fig. 8(a), the region of reduced flow density is distinctly sepa- 98
rated from the deflected OSW, which is very different from the
33 (1) The mechanism of formation and physical characteristics 99
case in Fig. 8(b), where the region of reduced density follows
34 of ODWs over dual-angle wedges, with respect to the position of 100
closely behind the deflected OSW.
35 wedge deflection and the second wedge-angle, was investigated in 101
Fig. 9 further quantifies these phenomena with the help of the
36 this study. 102
EC

values of the OSW and ODW angles, and the shape of the OSW-
37 (2) Analytical modeling of one-dimensional flow conservation 103
ODW transition. For deflection at P.2, in comparison to the theo-
38 equations with finite-rate chemical reaction can predict ODW in- 104
39 retical OSW angle of 19.3◦ for a 12◦ wedge, the deflected OSW duction and initiation lengths in dual angle-wedge with reasonable 105
40 angle gradually increases from 22◦ to 25◦ prior to a smooth tran- accuracy. 106
41 sition. While for the same second wedge-angle of 12◦ in Fig. 7(a) (3) Two kinds of detonation formation mechanisms were ob- 107
for deflection at P.1, the OSW angle remains largely steady, close
RR

42 served over the dual-angle wedges. The first one is an abrupt 108
43 to 21◦ prior to an abrupt transition. The initial drop seen in the transition from an OSW into an ODW, as commonly observed for 109
44 shock angle for the first case (P.1) is possibly because the OSW is planar ODWs. This mechanism was observed for the 21◦ –12◦ and 110
45 not fully resolved at the beginning of the wedge and the expansion 21◦ –15◦ dual-angle wedges deflected at P.1. Whereas, the second 111
46 takes place relatively early before a stable shock angle is attained. mechanism involves a brief induction process followed by a span 112
47 These observations for the 21◦ –12◦ dual-angle wedge makes an of shock-induced combustion that leads to an ODW over the sec- 113
CO

48 interesting case in the shock-to-detonation transition phenomena, ond wedge-angle. This mechanism was observed for the 21◦ –12◦ 114
49 where the higher wave angle for deflection at P.2 implies that the dual-angle wedge deflected at P.2. 115
50 deflected OSW is affected by the downstream heat-release. It is (4) For a given second wedge-angle, by inclining the wedge at 116
51 an example of a quenched shock-to-detonation transition as a re- a high first angle over only a short span, the induction length can 117
52 sult of the interaction between a significantly progressed induction be reduced significantly, along with a reduction in ODW initiation 118
53 process and the expansion waves over the wedge deflection re- length. 119
UN

54 gion. The premature wedge deflection leads to an “over-expanded” (5) The position of wedge deflection strongly determines the 120
55 shock-to-detonation transition process. Therefore, the combustion mechanism of shock-to-detonation transition and ODW initiation 121
56 phenomena behind the deflected OSW is a case of a very weak characteristics. When the deflection position is prior to induction 122
57 detonation, or characteristically—a shock-induced combustion. length value for the first wedge-angle, abrupt shock to detona- 123
58 The shock-to-detonation transition in Fig. 8(b) is further de- tion is observed with ODW initiation taking place over the second 124
59 layed by the intermediate shock-induced combustion. As a result, wedge-angle, with ODW properties primarily determined by the 125
60 the OSW curves upwards steadily from 22◦ to 25◦ (see Fig. 9), second wedge-angle. Where as if deflected close to the induction 126
61 and transition occurs after 40 mm along the horizontal axis of length for the first wedge angle, the ODW formation mechanism 127
62 the domain. The results in Fig. 7, Fig. 8, and Fig. 9 show that the is accompanied by a span of shock-induced combustion. Finally, 128
63 shock-to-detonation transition for the case is highly smooth, which if deflection position is moved further downstream, the induction 129
64 is expected from the high-temperature combustion region behind region shifts over to the first wedge-angle, until the ODW initia- 130
65 the deflected OSW. Transition from the upward turning OSW into tion takes place entirely over the first wedge-angle. Therefore, the 131
66 an ODW is very gradual that takes place over a span of 20 mm. wedge deflection position is a critical geometric parameter for the 132
 JID:AESCTE AID:4689 /FLA [m5G; v1.243; Prn:26/09/2018; 16:49] P.8 (1-8)
8 B. Bomjan et al. / Aerospace Science and Technology ••• (••••) •••–•••

1 determination and control of the formation characteristics of an [19] M.Y. Gui, B.C. Fan, G. Dong, Periodic oscillation and fine structure of wedge- 67
2 ODW. induced oblique detonation waves, Acta Mech. Sin. 27 (2011) 922–928. 68
3 [20] C.A.R. Pimentel, J.L.F. Azevedo, L.F. Figueira da Silva, et al., Numerical study 69
of wedge supported oblique shock wave-oblique detonation wave transitions,
4 Conflict of interest statement J. Braz. Soc. Mech. Sci. XXIV (2002) 149–157.
70
5 [21] P.M. Rubins, T.H.M. Cunningham, Shock-induced supersonic combustion in a 71
6 None declared. constant-area duct, J. Spacecr. Rockets 2 (1965) 199–205. 72
7 [22] S.A. Ashford, G. Emanuel, Wave angle for oblique detonation waves, Shock 73
Waves 3 (1994) 327–329.
8 References [23] J.A. Fort, A Numerical Study of Attached Oblique Detonation, PhD diss., Univer-
74
9 75
sity of Washington, 1993.
10 [1] D.T. Pratt, J.W. Humphrey, D.E. Glenn, Morphology of standing oblique detona- [24] C.I. Morris, Shock-Induced Combustion in High-Speed Wedge Flows, PhD diss., 76
11 tion waves, J. Propuls. Power 7 (1991) 837–845. Stanford University, 2001. 77
[2] S.A. Ashford, Oblique Detonation Waves, with Application to Oblique Detoaton [25] G. Fusina, J.P. Sislian, B. Parent, Formation and stability of near Chapman–
12 78
Wave Engines, and Comparison of Hypersonic Propulsion Engines, PhD diss., Jouguet standing oblique detonation waves, AIAA J. 43 (2005) 1591–1604.

OF
13 The University of Oklahoma, 1994. 79
[26] D.C. Alexander, J.P. Sislian, B. Parent, Hypervelocity fuel/air mixing in mixed-
14 [3] C. Viguier, A. Gourara, D. Desbordes, H2 -air and CH4 -air detonations and compression inlets of Shcramjets, AIAA J. 44 (2006) 2145–2155. 80
15 combustions behind oblique shock waves, Proc. Combust. Inst. 27 (1998) [27] D.C. Alexander, J.P. Sislian, A computational study of the propulsive character- 81
2207–2214. istics of a Shcramjet engine, J. Propuls. Power 24 (2008) 34–44.
16 82
[4] D. Desbordes, L. Hamada, C. Guerraud, Supersonic H2 -air combustion behind [28] J. Chan, Numerically Simulated Comparative Performance of a Scramjet and
17 83
oblique shock waves, Shock Waves 4 (1995) 339–345. Shcramjet at Mach 11, Masters Thesis, University of Toronto, 2010.
18 [5] H.H. Teng, H.D. Ng, Z.L. Jiang, Initiation characteristics of wedge-induced 84
[29] J.P. Sislian, B. Parent, Hypervelocity fuel/air mixing in a Shcramjet inlet,

RO
19 oblique detonation waves in a stoichiometric hydrogen-air mixture, Proc. Com- J. Propuls. Power 20 (2004) 263–272. 85
20 bust. Inst. 36 (2017) 2735–2742. [30] S. Bhattrai, H. Tang, Formation of near-Chapman–Jouguet oblique detonation 86
[6] H.H. Teng, P.F. Yang, Z.L. Jiang, Numerical study of oblique detonation initiations wave over dual-angle ramp, Aerosp. Sci. Technol. 63 (2017) 1–8.
21 87
with chain branching kinetics, AIAA SciTech Forum, 2017. [31] Y.N. Zhang, J.S. Gong, T. Wang, Numerical study on initiation of oblique deto-
22 [7] V.V. Vlasenko, V.A. Sabelnikov, Numerical simulation of inviscid flows with hy- 88
nations in hydrogen–air mixtures with various equivalence ratios, Aerosp. Sci.
23 drogen combustion behind shock waves and in detonation waves, Combust. Technol. 49 (2015) 130–134. 89
24 Explos. Shock Waves 31 (1995) 376–389. [32] Y. Fang, et al., Numerical study of inflow equivalence ratio inhomogeneity on 90

DP
[8] H.H. Teng, W. Zhao, Z.L. Jiang, A novel oblique detonation structure and its oblique detonation formation in hydrogen-air mixtures, Aerosp. Sci. Technol. 71 91
25
stability, Chin. Phys. Lett. 24 (2007) 1985–1988. (2017) 256–263, https://doi.org/10.1016/j.ast.2017.09.027.
26 92
[9] H.H. Teng, Z.L. Jiang, On the transition pattern of the oblique detonation struc- [33] P. Boivin, Reduced-Kinetic Mechanisms for Hydrogen and Syngas Combustion
27 ture, J. Fluid Mech. 713 (2012) 659–669. 93
Including Autoignition, PhD diss., University of Malaga, 2011.
28 [10] L.F. Figueira da Silva, B. Deshaies, Stabilization of an oblique detonation [34] P. Saxena, F.A. Williams, Testing a small detailed chemical-kinetic mechanism 94
29 wave by a wedge: a parametric numerical study, Combust. Flame 121 (2000) for the combustion of hydrogen and carbon monoxide, Combust. Flame 145 95
152–166. (2006) 316–323.
30 96
[11] Y. Liu, Y.S. Liu, D. Wu, J.P. Wang, Structure of an oblique detonation wave in- [35] C. Li, K. Kailasanath, E.S. Oran, Effects of Boundary Layers on Oblique Detona-
31 duced by a wedge, Shock Waves 26 (2016) 161–168. 97
TE
tion Structures, AIAA paper 93-0450, 1992.
32 [12] Y. Liu, X.D. Han, S.B. Yao, J.P. Wang, A numerical investigation of the prompt [36] M.R. Kamel, C.I. Morris, I.G. Stouklov, et al., PLIF imaging of hypersonic reac- 98
33 oblique detonation wave sustained by a finite wedge, Shock Waves 26 (2016) tive flow around blunt bodies, in: Twenty-Sixth Symp. Combust., Combust. Inst, 99
34 729–739. Pittsburgh, 1996, pp. 2909–2915. 100
[13] J.Y. Choi, D.W. Kim, I.S. Jeung, et al., Cell-like structure of unstable oblique det- [37] Y. Daimon, A. Matsuo, Analogy between wedge-induced steady oblique detona-
35 101
onation wave from high resolution numerical simulation, Proc. Combust. Inst. tion and one-dimensional piston-supported unsteady detonation, Sci. Technol.
36 31 (2007) 2473–2480. 102
EC

Energ. Mater. 65 (2004) 111–115.

37 [14] M.A.T. Walter, L.F. Figueira da Silva, Numerical study of detonation stabilization [38] Y. Huang, H. Ji, F.S. Lien, H. Tang, Three-dimensional parallel simulation 103
38 by finite length wedges, AIAA J. 44 (2006) 353–361. of spinning detonation in narrow square tube, Chin. Phys. Lett. 29 (2012), 104
[15] H.H. Teng, Z.L. Jiang, H.D. Ng, Numerical study on unstable surfaces of oblique 114701(1-4).
39 105
detonation, J. Fluid Mech. 744 (2014) 111–128. [39] A.S. Durna, M.E.H.A. Barada, B. Celik, Shock interaction mechanisms on a dou-
40 [16] P.F. Yang, H.D. Ng, H.H. Teng, Z.L. Jiang, Initiation structure of oblique deto- 106
ble wedge at Mach 7, Phys. Fluids 28 (2016), 096101(1-17).
41 nation waves behind conical shocks, Phys. Fluids 29 (2017) 086104, https:// [40] M.S. Liou, A sequel to AUSM: AUSM+, J. Comput. Phys. 129 (1996) 364–382. 107
RR

42 doi.org/10.1063/1.4999482 (1-10). [41] M.O. Connaire, H.J. Curran, J.M. Simmie, et al., A comprehensive modeling study 108
43 [17] J. Verreault, A.J. Higgins, R.A. Stowe, Formation of transverse waves in oblique of hydrogen oxidation, Int. J. Chem. Kinet. 36 (2004) 603–622. 109
detonations, Proc. Combust. Inst. 34 (2013) 1913–1920.
44 110
[18] J. Verreault, A.J. Higgins, Formation and structure of steady oblique and conical
45 detonation waves, AIAA J. 50 (2012) 1766–1772. 111
46 112
47 113
CO

48 114
49 115
50 116
51 117
52 118
53 119
UN

54 120
55 121
56 122
57 123
58 124
59 125
60 126
61 127
62 128
63 129
64 130
65 131
66 132