Comparison of wing–propeller interaction in tractor and pusher configuration

Abstract

This research compared a tractor and a pusher configuration for a tilt-body vertical take-off and landing micro air vehicle using an experiment in a subsonic wind tunnel. All tests were conducted in the range 6–10 m/s and 6000–8000 RPM of the freestream and the propeller rotation, respectively. The wing model was a rectangular platform with an NACA 0012 airfoil and AR = 1. The incidence angles varied within the range 0°–90° to cover flight configuration from cruise to hover and vice versa. Basically, the downstream of the propeller is stronger than the upstream. Thus the wing submerses in the propeller’s downstream, which is the tractor configuration, and has an advantage over the other configuration by increasing the freestream velocity and decreasing the angle of attack. The results of the experiment found that the wing aerodynamics of the tractor were improved by the prop-wash effect with increases of about 1–1.13 times in the wing lift curve slope and about 0.9–1.12 times in the factor K of the drag polar curve. However, in the case of the pusher, the change was not significant. The stall angle, the maximum lift coefficient, and the drag coefficient at zero angle of the wing all increased in both the tractor and the pusher configurations due to the prop-wash effect. Moreover, after experiencing the prop-wash effect, the aerodynamic center of the tractor wing was almost at the same point.

Keywords

Micro Air Vehicle Prop-wash Effect Wing-Propeller Interaction

Introduction

Nowadays, many configurations of tilt-body micro air vehicles (MAVs) have been developed, for example, fixed wing and flapping wing. The fixed wing was the focus of this research with the propeller’s position being the specific area of study. Normally, there are two main types of propeller position; where the propeller is mounted at the wing leading edge and is known as a “tractor” and where the propeller is mounted at the wing tailing edge, which is known as a “pusher.” The wing aerodynamic characteristics of each are different and their force and moment parameters were compared in this study. Moreover, the interaction between the wing and propeller in the type of pusher configuration was studied in this research. Tractor configuration has been formerly studied by Chinwicharnam et al.¹

There are several MAV tractor configurations such as Mini-Vertigo and MAVion shown in Figure 1(a) and (b), respectively. The Mini-Vertigo MAV, which was developed by Randall and Shkarayev^2,3 and Bataille,⁴ is a coaxial propeller mounted at the wing leading edge. These studies were performed in a subsonic wind tunnel, and the results showed that the slipstream flow influences the stall delay, lift enhancement and drag and can increase the aerodynamic efficiency. The MAVion from Supaero with two propellers was designed by Itasse.⁵ The propeller’s dimension of the MAVion covers almost the entire wing span, which consequently improves the wing aerodynamic characteristics due to the propeller downstream flowing around the wing. Another type of research of the wing and propeller interaction in the tractor configuration was studied by Deng et al.⁶ using both experimental and numerical methods. They found that the slipstream has a significant influence on the pressure distribution on the wing surface, as well as investigating and explaining the pattern of the wing-tip vortex at different angles of attack with a rotary propeller.

Figure 1.

Current micro air vehicles: (a) Mini-Vertigo, (b) MAVion, and (c) Disk-wing aircraft.

Besides the tractor configuration, there is a new conceptual design for a disk-wing aircraft shown in Figure 1(c) which has been recently reported by Ageev.⁷ This aircraft was designed using a slot in the middle of the wing for a propeller in order to take advantage of both the propeller upstream and the propeller downstream. His results showed the good stall characteristics of this concept, a low lift-to-drag ratio and a sophisticated flow pattern. However, this configuration has only been reported in a few investigations.

The pusher configuration is not as well known as the tractor configuration. Work was carried out by Choi and Ahn⁸ using a commercial computational fluid dynamics tool, FLUENT. They found that the propeller effect retarded the flow separation on the upper surface by 4° and it promoted the reattachment of the separated flow, which contributed to increasing the lift and the pitching-up moment on the vehicle, compared with the wing without a propeller. The lift-to-drag ratio was observed to be slightly lowered by using a propeller, contrary to the commonly known benefits of suppressed separation.

The propeller’s position is very important for the improvement of wing aerodynamic characteristics as studied by Catalano,⁹ Hrishikeshavan,¹⁰ Shkarayev,¹¹ and Veldhuis¹² for a full scale aircraft; the position of the propeller has an influence on the wing boundary layer characteristics such as laminar flow extension and transition, laminar separation bubbles, and reattachment and turbulent separation. Additionally, the propeller slipstream produced a modified flow over the wing where the kinetic energy increased, the Reynolds number at wing surface increased and the separation behavior changed. Therefore, the interaction between the wing and propeller in a comparison of the tractor and pusher configurations was investigated in the current research.

In this study, the difference between the aerodynamic characteristics of the tractor and pusher MAV configurations was investigated using six types of models: the wing without and with propeller in pusher and tractor configurations; a separated wing and propeller in both the tractor and pusher; and the last model involved propeller testing which can be seen in more detail later in this paper. Testing measured the aerodynamic forces and the moment of the models using very accurate facilities and the methodology of a wind tunnel experiment as suggested by Pope.¹³ Furthermore, all the models were tested to see the effect of propeller wash on the wing aerodynamics during transition flight. Moreover, the wing downwash effect over the propeller, which several researches had neglected, was considered in this paper.

Model test

The first campaign has already been completed using the SabRe wind tunnel at Supaero, Toulouse, France. The model test was a tractor configuration and the case study was published in the literature.¹ The current campaign was performed using both the tractor configuration and the pusher configuration in order to compare a tractor and a pusher wing. The specifications of the current model are shown in Table 1 and Figure 2(a) and (b), which are similar to the previous model, but they were built using a different 3D printer machine.

Figure 2.

Models set up in the wind tunnel: (a) tractor and (b) pusher.

Table 1.

Model information.

Wing	Information	Value
	Airfoil	NACA 0012
	Wing chord	0.3 m
	Wing Span	0.3 m
	Wing area	0.09 m²
	Wing platform	Rectangular
	Material	PLA (Polylactide)
Motor	Type	Brushless
	Brand	ALBATROSS 2215–1700 KV
	Weight	70 g
Blade	Type	Garupner super nylon
	Size	8 × 6″
	Blade	2

The wing model is a symmetrical airfoil that has the wing lift at zero angle of attack and the aerodynamic center is at about 25% of a wing chord. The tractor has the propeller located far from the wing leading edge at 0.042 m and also the pusher has the propeller located far from the trailing edge at 0.042 m. The wing surface was made quite smooth using some sandpaper and by painting. These processes were delicately undertaken because of concerns about affecting the wing shape.

Experimental setup

A subsonic wind tunnel at Kasetsart University (KU), Bangkok, Thailand was used in this experiment as shown in Figure 3. The wind tunnel is a closed loop and generates an airspeed in the range 6–10 m/s. The experiments were carried out in the square cross-section which has dimensions of 1 m × 1 m × 3 m (width × height × length) with a contraction ratio of 4.0. The fan speed was controlled by the container control panels. The freestream velocity was recorded using NI9215 hardware consisting of a National Instrument Data Acquisition system (NI-DAQ) controlled by the Labview software. The model was set in the middle of the test section and was connected to a test bench which contained a new, three-components, aerodynamic force balance. The force balance was mounted with a manual rotation system for increasing/decreasing the incidence angle. The balance shown in Figure 4 was designed to measure the lift, drag, and pitching moment simultaneously. Three load cells were installed on the balance and it was connected to the NI9237 for sample recording at 10 kHz. All load cells were calibrated using some standard weights.

Figure 3.

Subsonic wind tunnel at Kasetsart University with the experiment facilities.

Figure 4.

Three-components force balance.

The installation scheme of the test bench mechanism at the testing section is shown in Figure 5(a). The level adjustment was carefully checked parallel to the wing chord line and the freestream direction. Six experiments were designed in this study as shown in Figure 5(b)–(g): propeller test (PT), wing test (WT), mounted propeller on wing model of the tractor/pusher test (MPWT/MPWP), and separated propeller from wing model in configurations of the tractor/pusher (SPWT/SPWP), consecutively.

Figure 5.

Experiment setup of (a) components of the test bench and (b)–(g) six experiments.

Test descriptions data accuracy

The wing (WT) model was tested and data were recorded on the lift force, drag force, and pitching moment as illustrated in Figure 6(a)–(c). The results recorded at Supaero were used for comparison to investigate data accuracy. Both the wind tunnel tests produced the same aerodynamic behavior of the forces and moment for various angles of attack. The lift coefficient at Supaero was a little smaller than for the results at KU but was not great taking into account the error bars. Moreover, the stall of both was recorded at two angles of 25° and 45°, and the curve of the drag force and pitching moment coincided well with the results from Supaero. The aerodynamic forces and moments that were plotted with various Reynolds numbers from 129,000 to 300,000 based on the wing chord showed that the results of the forces and moment were the same curve by changing the angle of attack, e.g., the lift curve slope, stall angle, C_D₀, and pitching-up or -down angles. It can be said that there is small effect of the Reynolds number for the low aspect ratio wing.

Figure 6.

Wing aerodynamic characteristics of forces and moment versus $α_{MAV}$ .

The aerodynamic characteristics of an isolated propeller alone (PT) were recorded in the dynamic test which considered the freestream at 10 and 6 m/s, and the propeller speed at 4000–6000 RPM as shown in Figure 7(a) and (b). The thrust force was recorded at various angles of attack from 0° to 90° in order to observe the propeller aerodynamic behavior. It was found that the thrust force increases as both the propeller speed and the angle of attack increase. This can be explained by the angles of attack increasing while the freestream velocity of the propeller in the axial direction or the axial velocity is reduced. Note that the behavior of a propeller at 90° with airspeed is like the propeller at 0° without airspeed due to the fact that the axial velocity of a propeller at 90° with airspeed is equal to zero. The thrust coefficient measured at Supaero was compared in Figure 7(b) with a propeller speed of 6000 and V = 6 m/s. There was a small error between the results of Supaero and KU that was considered acceptable.

Figure 7.

Propeller aerodynamic characteristics of forces and moments versus α.

The mathematical model of the propeller thrust coefficient as a function of the advance ratio and angle of attack was determined and is expressed in equation (1) when V = 10 m/s, and in equation (2) when V = 6 m/s. The advance ratio is relative to the freestream velocity, propeller speed, and propeller diameter as shown in Table 2.

C_{T} ≅ 5.6213 J^{2} - 8.1388 J + 0.00004 α_{MAV}^{2} + 0.001 α_{MAV} + 2.846

(1)

C_{T} ≅ 0.0614 J^{- 2.463} + 0.00004 α_{MAV}^{2} + 0.003 α_{MAV} + 0.6739

(2)

Table 2.

Advance ratio equivalences.

V(m/s)	Re	RPM	J
6	12,9000	6000	0.300
6	12,9000	8000	0.225
10	215,000	6000	0.500
10	215,000	8000	0.375

Free body diagram of models

The free body diagram of the models, a wing, propeller, tractor, and pusher, are shown in Figure 8(a)–(d), respectively. The isolated wing generates forces of lift (L), drag (D), and moment (M) which are a vector operation by a normal force (N) and axial force (A). A propeller usually creates thrust force (T), lateral force (Np), and torque (Q). The propeller moment can be considered to be zero at the center of the propeller. Furthermore, the propeller generates an induced velocity (w) in a downstream tube. Likewise, the tractor and pusher configurations have the component of the forces and moment as the combination of the wing and the propeller which has been previously mentioned.

Figure 8.

Forces and moments diagrams.

This research considered in particular the relationships of the freestream velocity and the propeller rotation in terms of the advance ratio which are shown in Table 2. The formula equations for the aerodynamic coefficients, Reynolds number, and advance ratio in this paper can be calculated as

C_{L} = \frac{L}{\frac{1}{2} ρ V^{2} S}, C_{D} = \frac{D}{\frac{1}{2} ρ V^{2} S}, C_{M} = \frac{L}{\frac{1}{2} ρ V^{2} Sc}, C_{X} = \frac{D}{\frac{1}{2} ρ V^{2} S}, C_{T} = \frac{T}{\frac{1}{2} ρ V^{2} S}, C_{Np} = \frac{Np}{\frac{1}{2} ρ V^{2} S}, Re = \frac{ρ Vd}{μ}, J = \frac{V}{nd}

Results

Tractor/pusher wing and wing comparison

Lift coefficient

The aerodynamic behavior of the lift coefficient as a function of the full range of angle of attack for the SPWT (Separated Propeller-Wing in Tractor Configuration), SPWP (Separated Propeller-Wing in Pusher Configuration), and WT configurations is compared in Figure 9(a)–(d) with different advance ratios from 0.225 to 0.5. The lift coefficient of all wings clearly increased as the angles of attack increased and it increased until the stall angle. Then after stalling, it slightly declined due to the high separation the adverse pressure gradient presented to the flow. Notably, the stall angle of the tractor/pusher wing was delayed from 20° to a range between 45° and 55° by the effect of prop-wash. Furthermore, the prop-wash effect increased the maximum lift coefficient (C_Lmax) and the lift–curve slope (C_L_α) in both the tractor/pusher wings.

Figure 9.

Comparison of lift coefficient of the wing prop-off and prop-wash effect versus $α_{MAV}$ .

The results shown in Figure 9(a)–(d) illustrate that the performance of the tractor wing is clearly better than for the pusher wing profile. The performance of both the tractor and pusher wing profiles is characterized by the lift coefficient. The results showed that the C_Lmax of the tractor wing increased from the C_Lmax of the wing alone in the range 91–232%, while the C_Lmax of the pusher wing increased only 76–104% for the wing alone. This was due to the fact that the wing of the tractor benefits from the increasing airspeed emanating from the propeller downstream (V + w), while the wing of the pusher benefits from the propeller upstream. It should be noted that the propeller downstream had a higher flow energy instance (i.e., a turbulent and complex flow) than the propeller upstream. Moreover, the wings will have an effective angle of attack as mentioned in the study by Chinwicharnam et al.¹ It seems the tractor wing has a less effective angle (α _wing ) than the wing pusher and has a new, higher airspeed (V′). The values for C_Lmax for all wings are shown in Table 3.

Table 3.

C_L_{α (0–15°)} and C_Lmax for wing, tractor, and pusher configurations with their corresponding angle of attack.

J	Wing			Tractor			Pusher
J	C_Lmax	α _stall	C_L _α(0–15°)	C_Lmax	α _stall	C_L _α(0–15°)	C_Lmax	α _stall	C_L _α(0–15°)
0.225	0.876 (Re 129,000)	25°	0.039	2.908	55°	0.062	1.786	50°	0.041
0.300	0.876 (Re 129,000)	25°	0.039	2.430	50°	0.061	1.657	45°	0.040
0.375	0.855 (Re 215,000)	25°	0.039	1.960	50°	0.047	1.629	50°	0.040
0.500	0.855 (Re 215,000)	25°	0.039	1.634	45°	0.044	1.505	45°	0.039

When considering the difference between the C_Lmax of the tractor wing and the pusher wing at J = 0.225, 0.3, 0.375, and 0.5, the C_Lmax of tractor wing was greater by around 128%, 88%, 39%, and 15%, respectively, as shown in Figure 9(a)–(d). The difference in C_Lmax reduced when the advance ratio increased because the lift coefficient is in inverse proportion to the freestream (V) and the wing lift coefficient of the pusher is quite independent of the airspeed or the advance ratio.

Apart from the lift–curve slope (C_L_α) of the wing alone, the angle of attack from 0° to 15°, was constant at 0.039 for each advance ratio. The C_L_α values of the tractor wing were 0.062, 0.061, 0.047, and 0.044, respectively, for advance ratios of J = 0.225, 0.3, 0.375, and 0.5, and were 0.041, 0.040, 0.040, and 0.039, respectively, for the pusher wing as shown in Table 3. In general, the C_L_α of the tractor wing was higher than for the pusher wing at every advance ratio due to the fact that the wing of the tractor has a propeller boost airspeed that promotes an increase in the lift force.

The advance ratio was increased by decreasing the propeller rotational speed, and this is shown in the comparison between Figure 9(a) and (b), and between Figure 9(c) and (d). It was found that the C_L_α of both the tractor/pusher wings was reduced as illustrated in Table 3 because the induced velocity of the propeller (w) dropped due to the decreasing slipstream resulting velocity (V_R). The slipstream resulting velocity of the tractor wing was plotted with model angles at various advance ratios as shown in Figure 10, according to calculation using the McCormick¹⁴ equation. Note that the propeller upstream of the pusher wing did not experience the induced airspeed, but its trend of slipstream resulting velocity versus model angles should be similar to that of the tractor wing.

Figure 10.

Comparison of the propeller speed influence on the slipstream resulting velocity and freestream velocity ratio of tractor wing versus $α_{MAV}$ .

Drag coefficient

The aerodynamic behavior of the three particular models is explained through a comparison of the drag coefficient for advance ratios between 0.225 and 0.5, as illustrated in Figure 11(a)–(d). Even if the prop-wash can enhance the wing lift and even extend the stall, it will simultaneously induce drag force in both the tractor and pusher configurations. The drag coefficient of the wing with the propeller increases because the prop-wash will generate a complex flow with high turbulence and also increase the freestream velocity across the wings and the propeller. Indeed, the increasing freestream velocity will influence the wing tip vortex. Because the wing model used in this study is a low aspect ratio, the drag induced from the wing alone model is significantly contributed to by the tip vortex. The wing tip vortex increases as the angle of attack is increased. It therefore can be said that the drag of the wing increases as the angle of attack increases. For the tractor and pusher configurations, the drag coefficient increased until it reached the stall angle after which it then decreased. The maximum value of C_D depends upon the advance ratio. Table 4 shows the maximum C_D for the pusher and tractor configurations for four different advance ratios and their corresponding angle of attack.

Figure 11.

Comparison of drag coefficient of the wing prop-off and prop-wash effect versus $α_{MAV}$ .

Table 4.

Maximum C_D for pusher and tractor configurations and their corresponding angle of attack.

J	Wing		Tractor		Pusher
J	C_D( _max)	α _{CD (} _max)	C_{D (} _max)	α _{CD (} _max)	C_{D (} _max)	α _{CD (} _max)
0.225	1.56 (Re 129,000)	90°	4.25	70°	2.8	65°
0.300	1.56 (Re 129,000)	90°	3.20	60°	2.5	60°
0.375	1.48 (Re 215,000)	90°	2.20	55°	2.25	60°
0.500	1.48 (Re 215,000)	90°	1.80	50°	1.8	60°

It can be seen from the results accumulated in Table 4 that for both the pusher and tractor configurations, the C_Dmax decreases as the advance ratio increases. Figures 11(a) and 7(b), respectively, represent the drag coefficient results for J = 0.225 and J = 0.3 and show that the drag coefficient of the tractor wing is always higher than for the pusher wing because the tractor wing submerges in the propeller downstream. Ideally, the propeller downstream generates a turbulent flow to the wing where this turbulent flow is also stronger than the wing submerged in the propeller upstream. Therefore, the tractor wing increases the parasite drag, defined as the summation of skin friction and pressure drag. For advance ratios of 0.375 and 0.5 shown in Figure 11(c) and (d) which correspond to a velocity of 10 m/s, the drag coefficient of the tractor wing is a little lower than the pusher wing profile. This might be caused by the decrease in the induced velocity of the propeller. When the propeller induced velocity (w) decreased due to the increasing airspeed (V = 10 m/s), the new angle of attack (α _w ) of the wing due to the combination of airspeed and propeller-induced velocity (V + w) must be lower than the tractor wing at V = 6 m/s. The minimum drag coefficient of all models was found at 0° and the factor K of the drag polar curve which is used to calculate C_D of wing in equation (3)) is shown in Table 5.

C_{D} = C_{D 0} + {KC}_{L}^{2}

(3)

Table 5.

Minimum C_D and the factor K of the drag polar equation for pusher and tractor configurations.

J	Wing		Tractor		Pusher
J	C_D ₀	K (0°–20°)	C_D ₀	K (0°–20°)	C_D ₀	K (0°–20°)
0.225	0.014 (Re 129,000)	0.45	0.099	0.28	0.088	0.52
0.300	0.014 (Re 129,000)	0.45	0.093	0.29	0.115	0.48
0.375	0.014 (Re 215,000)	0.45	0.091	0.30	0.102	0.44
0.500	0.014 (Re 215,000)	0.45	0.070	0.39	0.101	0.44

Aerodynamic efficiency

The lift-to-drag ratio, which can be represented as the efficiency of MAVs, was plotted by changing the advance ratios from 0.225 to 0.5 for various angles of attack as shown in Figure 12(a)–(d). The results of all models have a general trend which increased greatly over a small range of angle (0°–15°) and then reached a maximum L/D ratio after which the trend gradually reduced. The maximum L/D ratio of the wing with the propeller in both tractor and pusher configurations was lower than for the wing alone because the prop-wash effect induced a drag force to the wing. Considering the advance ratios of J = 0.225, 0.3, 0.375, and 0.5, the maximum L/D ratio for the tractor wing decreased from the maximum L/D ratio of wing without a propeller by 15%, 30%, 25%, and 30%, respectively, and also the maximum L/D ratio for the pusher wing decreased 28% 35%, 41%, and 40%, respectively, of the maximum L/D ratio of the wing without a propeller. The trend of L/D at each advance ratio can be divided into two zones—wing pre-stall and wing post-stall. In the wing pre-stall zone, the L/D ratio of the wing decreased when the propeller was attached and the L/D ratio of the pusher wing was the lowest. In the wing post-stall zone, the tractor wing had the highest aerodynamic efficiency due to the fact that the prop-wash effect improves the flow around the tractor wing, such as in a delayed stall, increasing the airspeed and promoting flow reattachment. The highest aerodynamic efficiency of the tractor wing is very useful for the tilt-body MAV configuration because it will save battery power during the flight and be an advantage during a transition flight. The MAV can easily tilt its body from horizontal flight to vertical flight using a small amount of battery energy. Moreover, it increases the performance of the MAV at a very high incidence angle.

Figure 12.

Comparison of aerodynamic efficiency of the wing prop-off and prop-wash effect versus $α_{MAV}$ .

Table 6 shows the maximum L/D ratio for the wing, tractor, and pusher configurations for four different advance ratios and their corresponding angles of attack. The maximum L/D ratio of the wing without a propeller was at 10°, and it increased as the Reynolds number increased. The maximum L/D ratio of the tractor wing decreased when the advance ratio was increased by decreasing the RPM only; for example, J = 0.225 to J = 0.3. The maximum L/D ratio of the tractor wing was changed a small amount by the advance ratio. It should be noted that both the tractor and pusher wings had an angle of maximum L/D ratio of around 10°–15° which is dependent on the advance ratio.

Table 6.

Maximum ratio (L/D)_max for wing, tractor, and pusher configurations and their corresponding angles of attack.

J	Wing		Tractor		Pusher
J	(L/D) _max	α₍ _L _/ _D ₎ _max	(L/D) _max	α₍ _L _/ _D ₎ _max	(L/D) _max	α₍ _L _/ _D ₎ _max
0.225	3.50 (Re 129,000)	10°	2.97	10°	2.52	15°
0.300	3.50 (Re 129,000)	10°	2.47	15°	2.27	15°
0.375	4.16 (Re 215,000)	10°	3.14	15°	2.45	10°
0.500	4.16 (Re 215,000)	10°	2.55	15°	2.48	10°

Moment coefficient

A part of the wing moment behavior, for example pitching up/down, is also important in the design and development of MAVs. The pitching moment of the isolated wing and the wing with propeller were measured at 25% of the wing chord-wise for a full range of angles of attack as shown in Figure 13(a)–(d). All models generated a negative moment or pitching up in the range of angle of attack from 0°–10° or 0°–15°, depending on their advance ratio. Above this angle, the pitch slope of all wings rose with a further increasing angle of attack up to stalling. At an angle of 20°–40° in every advance ratio, the positive moment of the tractor wing was less than the wing alone, due to the fact that the prop-wash effect improved the flow around the tractor wing. For example, it decreased the angle of attack, increased the airspeed, and delayed stalling. Above about 40°, the pitching-up moment of the tractor wing still increased and was higher than for the wing alone. Increased pitching up is generated by the wing lift, and this contributes to the prop-wash effect. However, the pusher wing had the highest pitching up and the lowest pitching down for every angle of attack and every advance ratio. The moment of the tractor wing was more stable than for the pusher wing at every angle, although the pitching down of the tractor wing was lower at a small angle of attack, but it was just a small value.

Figure 13.

Comparison of pitching moment coefficient of the wing prop-off and prop-wash effect versus $α_{MAV}$ .

According to the plot of C_M for various angles of attack, the slope of the moment coefficient (C_m,_{α (0–5°)}) was found by using an angle from 0° to 5°, and the aerodynamic center was calculated as a percentage of the chord as shown in Table 7. The C_m,_α(0–5°) of the wing was reduced by increasing the Reynolds number, while the C_m,_α(0–5°) of the tractor was not very sensitive to the various advance ratios. On other hand, the C_m,_α(0–5°) of the pusher increased by increasing the advance ratio. The aerodynamic center of both the isolated wing and the tractor was extended to 30% of the chord by increasing the freestream, but in case of the pusher, there was not much change.

Table 7.

C_m,_{α (0–5°)} and ac. for wing, tractor, and pusher configurations (Boldface in Table 7 shows a change or move of ac).

J	Wing		Tractor		Pusher
J	C_m, _{α (0–5°)}	ac. (%c)	C_m, _{α (0–5°)}	ac. (%c)	C_m, _{α (0–5°)}	ac. (%c)
0.225	−0.0010 (Re 129,000)	27.83	−0.0022	28.50	−0.0012	27.90
0.300	−0.0010 (Re 129,000)	27.83	−0.0023	28.86	−0.0010	27.60
0.375	−0.0018 (Re 215,000)	30.07	−0.0024	30.15	−0.0008	27.14
0.500	−0.0018 (Re 215,000)	30.07	−0.0023	30.28	−0.0007	26.87

Wing and prop-wash effect

Lift coefficient

The total lift coefficient of the pusher can be determined using equation (4) which combines the lift of the wing, the propeller, the effect of propeller wash (ΔC_Lprop→wing), and the effect of wing wash (ΔC_Lwing→prop). The propeller lift and wing-wash effects are considered as shown in Figure 14(a) and (b). The lift curve slope (C_L_α) of the total lift increased to 0.050 and 0.044 when the advance ratio was 0.375 and 0.5, respectively. It should be noted that the C_L_α values are calculated by using an angle range of 0° to 15°.The stall angle of the wing prop-on (Total) is at the same point as the wing with prop-wash effect. The propeller lift force increased with increasing angle of attack due to the decrease in the axial velocity of the propeller. When the propeller lift force is considered, the maximum of C_Ltotal at J = 0.375 increased 61% from the maximum wing prop-wash while at J = 0.5, the maximum C_Ltotal increased 35% from the maximum wing prop-wash.

C_{Ltotal} = C_{Lwing} + C_{Lprop} + Δ C_{Lprop \to wing} + Δ C_{Lwing \to prop}

(4)

Figure 14.

Wing prop-on/prop-off and prop-wash effect in terms of lift coefficient versus $α_{MAV}$ : (a) J = 0.375 and (b) J = 0.5.

A comparison between the prop-wash effect (ΔC_Lprop→wing) and the wing-wash effect (ΔC_Lwing→prop) is shown in Figure 15(a) and (b). At advance ratios of 0.375 and 0.5, the propeller wash had a small effect in the range 0°–15°; the wing has an effective angle which is not as high as the model angle, and at the same time, as this wing has a low aspect ratio, the lift curve is slight. Above an angle of 15°, the lift force increased with increasing angle of attack until the stall angle. Then, the lift coefficient gradually decreased for post-stall angles. A small wing-wash effect was observed compared with the prop-wash effect.

Figure 15.

Prop-wash and wing-wash effect in terms of lift coefficient versus $α_{MAV}$ : (a) J = 0.375 and (b) J = 0.5.

Total longitudinal force coefficient

The interaction between the pusher wing and the propeller in terms of the drag force coefficient and the total longitudinal force coefficient are shown in Figure 16(a) and (b). The negative C_D represents the thrust direction of models which can be found in the propeller force and the wing prop-on. It is clear that the curve of C_D of wing prop-on shifted down in quantify of the propeller C_D. Therefore, it is possible to derive equation (5) to explain this relation. The C_D_α ₌ ₀ of the wing prop-wash effect was higher than the wing prop-off due to the propeller increasing the freestream velocity and turbulent flow around the pusher wing. The propeller had a constant effect on the wing for the range 0°–20° which is the angle before wing prop-off stall; it can be observed that the difference between the C_D of the wing prop-off and the wing prop-wash was constant in this angle range because the effective angle of attack (α _w ) changed little when the wing increased the angle of attack (α _MAV ) and it happened only at an angle before the stall of wing prop-off. After an α _MAV value of 20°, the wing prop-wash strongly increased due to the complex flow by the prop-wash effect. It should be noted that the propeller can produce more thrust when the angle of attack is increased due to the fact that the freestream velocity, which is perpendicular to the propeller plane, is decreased by the increasing propeller angle of attack.

C_{Xtotal} = C_{Dwing} + C_{Dprop} + Δ C_{Dprop \to wing} + Δ C_{Dwing \to prop}

(5)

Figure 16.

Wing prop-on/off and prop-wash effect in terms of drag coefficient versus $α_{MAV}$ : (a) J = 0.375 and (b) J = 0.5.

The drag coefficients of the prop-wash effect (ΔC_Dprop→wing) and the wing-wash effect (ΔC_Dwing→prop) of the pusher wing for various angles of attack are shown in Figure 17(a) and (b). The propeller wash had a small effect on the wing in the range 0°–20° and then it increased with increment in the angle of attack. The drag force of the wing-wash effect was less than the prop-wash effect as the lift force. However, the results show that there was a minor effect of wing wash at J = 0.5.

Figure 17.

Prop-wash and wing-wash effect in terms of drag coefficient versus $α_{MAV}$ : (a) J = 0.375 and (b) J = 0.5.

Pitching moment coefficient

The pitching moment was measured at 25% of wing chord as shown in Figure 18(a) and (b), with a negative value representing pitching down and a positive value referring to pitching up. The total pitching moment can be determined using equation (6). The influence of C_Mprop was very slight compared with C_Mwing for various angles of attack due to the fact that there was only the C_Np influence on the propeller pitching moment. As mentioned in the previous studies,^1,15 the propeller at the incidence angle generated the resultant propeller thrust (C_T), which was not applied at the center of the propeller due to the asymmetric distribution of the thrust over the propeller disk. However, the radius of the current propeller was very small due to the fact that the total axial thrust does not displace more than 45% from the center of the propeller. Therefore, it is safe to assume that for this type of propeller, the pitching moment which is produced by C_Tr can be neglected.

C_{Mtotal} = C_{Mwing} + C_{Mprop} + Δ C_{Mprop \to wing} + Δ C_{Mwing \to prop}

(6)

where

C_{Mprop} = C_{T} r + C_{Np} x_{0.25 c}

Figure 18.

Wing prop-on/prop-off and prop-wash effect in terms of pitching moment coefficient versus $α_{MAV}$ : (a) J = 0.375 and (b) J = 0.5.

Comparison of the pitching moment coefficients of the prop-wash effect (ΔC_Mprop→wing) and the wing-wash effect (ΔC_Mwing→prop) are shown in Figure 19(a) and (b). The wing wash affected the pitching-down moment of the wing when J = 0.375 with quite a high result as shown in the plot. However, it was small compared with the pitching-up moment of the prop-wash effect in both advance ratios. The maximum C_M_(0.25 _c ₎ of the prop-wash effect occurred at 60° and 50°, when the advance ratio was 0.375 and 0.5, respectively.

Figure 19.

Prop-wash and wing-wash effect in terms of pitching moment coefficient versus $α_{MAV}$ : (a) J = 0.375 and (b) J = 0.5.

Total aerodynamic forces and moment of pusher

The mounted propeller on the wing model of the tractor/pusher test (MPWT/MPWP) was tested by measuring the lift force, drag force, and pitching moment as shown in Figure 20(a)–(f). The total forces and the moment are plotted to help users consider the capability of these models. The results were also plotted with the error bar from the experiment. The C_L_α values of both the tractor and pusher tended to increase when the advance ratio decreased, while the maximum C_L and stall angle increased as shown in Figure 20(a) and (b). The C_L of both tractor and pusher equaled zero, due to the fact that the wing had a symmetrical airfoil NACA 0012 and the Np of the propeller can be ignored. The longitudinal force of the tractor and pusher as shown in Figure 20(c) and (d) combined normally with the thrust and drag force. The tractor and pusher drag increased with an increasing angle of attack. The thrust of the low advance ratio was lower than the thrust of the high advance ratio because of the reduction in airspeed. Pitching-up or -down moment occurred with both the tractor and pusher as shown in Figure 20(e) and (f). The range of the pitching-down moment for the tractor was longer than for the pusher being almost 35° while the pusher was only 20°. The pitching-down moment at α _MAV = 0° of the pusher increased with a decrease in the advance ratio and was higher than for the tractor because the propeller lateral force (Np) of the pusher was located at the wing tailing edge as shown in Figure 8 and the length of x_0.25 _c of the pusher was more than for the tractor.

Figure 20.

Aerodynamic characteristics of tractor and pusher in various advance ratios: (a), (b) lift coefficient; (c), (d) drag coefficient; and (e), (f) pitching moment coefficient.

Conclusions

This research provides a comparison between the tractor-wing and the pusher-wing configuration of a tilt-body MAV using experimental data. Testing was performed using a low Reynolds number following the real-flight situation of MAVs at V = 6, 10 m/s and J = 0.225–0.5. The results were used to determine that the aerodynamic performance of the tractor configuration was better than for the pusher configuration, as the stall angle, lift–curve slope, maximum lift coefficient, and aerodynamic efficiency of the tractor wing were higher than for the pusher wing. Furthermore, the zero-lift drag coefficient of the tractor wing was lower than for the pusher wing. Thus, a tilt-body MAV can obtain great advantage from the tractor wing configuration; for example, a tractor wing during transition flight has a very high angle of attack and can generate a higher L/D ratio than a pusher wing. This will be useful in terms of saving battery power and increasing payloads.

The propeller improved the aerodynamic characteristics in both the wing tractor and pusher by increasing the lift and drag coefficient. It was noticed that the propeller increased the airspeed and decreased the angle of attack of the wing and helped the wing surface to have a greater reattachment boundary layer which extends the separation behavior and delays the wing-stall angle. These results showed that the tractor/pusher wing could take advantage from both the upstream and downstream of the propeller.

The aerodynamic total forces and moment of the pusher configuration were derived so that they combined the forces and moment of an isolated wing, an isolated propeller, a prop-wash effect, and a wing-wash effect. The prop-wash effect mainly increased the aerodynamic performance of the wing, but the wing–wash effect was just a small part of this interaction. However, the wing-wash effect cannot be ignored as also mentioned in previous study.¹ The total lift and drag coefficient, the maximum lift coefficient, the stall angle, the lift–curve slope, and the drag coefficient at zero angle were all higher as the advance ratio decreased.

The next configuration to be studied will be a mid-wing which has the propeller located in the wing body in a vertical direction. Perhaps, the performance of this configuration might be better than for either the tractor or pusher configuration because it benefits both the upstream and downstream of the propeller.

Footnotes

Acknowledgments

The authors would like to thank the students of Aerospace Engineering of Kasetsart University who helped us to perform the experimental set up.

Declaration of conflicting interests

The author(s) declared no potential conflict of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The authors would like to thank the Graduate School of Kasetsart University, Thailand. They appreciate their funding and budget support for this research.

Appendix

References

Chinwicharnam

Gomez

Mochetta

. Aerodynamic characteristics of a low aspect ratio wing and propeller interaction for a tilt-body MAV. Int J Micro Air Vehicle 2013; 5: 245–260.

Randall

Hoffmann

C-A

Shkarayev

. Longitudinal aerodynamics of a vertical take off and landing micro air vehicle. J Aircraft 2011; 48: 166–176.

Randall

Shkarayev

Abate

. Longitudinal aerodynamics of rapidly pitching fixed-wing micro air vehicle. J Aircraft 2012; 49: 453–467.

Bataille

Moschetta

Poinsot

. Development of a VTOL mini-UAV multi-tasking missions. Aeronaut J 2009; 113: 87–98.

Itasse

Moschetta

J-M

Ameho

. Equilibrium transition study for a hybrid MAV. Int J Micro Air Vehicles 2011; 3: 229–246.

Deng

van Oudheusden

Xiao

. A computational study on the aerodynamic influence of a propeller on an MAV by unstructured overset grid technique and low Mach number reconditioning. Open Aerosp Eng J 2012; 5: 11–21.

Ageev ND. Numerical investigation of disc-wing MAV with propeller in a wing slot. In: Proceedings of the international micro air vehicles conference, Multi-science, International Micro Air Vehicle conference and competitions 2011 (IMAV 2011), ‘t Harde, The Netherlands, 12–15 September 2011.

Choi S and Ahn J. A computational study on the aerodynamic influence of a pusher propeller on a MAV. In: Proceedings of 40th fluid dynamics conference and exhibit, Chicago, IL, 28 June–1 July 2010.

Catalano

. On the effects of an installed propeller slipstream on wing aerodynamic characteristics. Acta Polytech 2004; 44: 8–14.

10.

Hrishikeshavan V, Bogdanowicz C and Chopra I. Experimental investigation of performance of a wing-propeller system for a quad-rotor-biplane micro air vehicle. In: 54th AIAA/ASME/ASCE/AHS/ASC structures, structural dynamics and material conference, Boston, Massachusetts, 8–11 April 2013.

11.

Shkarayev

Hoffmann

. Longitudinal aerodynamics of a vertical takeoff and landing micro air vehicle. J Aircraft 2011; 48: 166–176.

12.

Veldhuis L. Propeller wing aerodynamic interference. The Netherlands: Delft University of Technology Netherlands, 2005.

13.

Pope

Barlow

Rae

. Low-speed wind tunnel testing, 3rd ed. New York: John Wiley & Sons, Inc., 2011.

14.

McCormick

. Aerodynamics of V/STOL flight, New York: Dover, 1967.

15.

Gomez D. Study of the sensitivity to the lateral wind of a mini unmanned aerial vehicle with VTOL flight capabilities. PhD Thesis, ISAE, France, 2013.