Experimental characterization of drag coefficient of an uav recovery parachute

Nội dung text: Experimental characterization of drag coefficient of an uav recovery parachute

1. JST: Smart Systems and Devices Volume 31, Issue 2, September 2021, 067-074 Experimental Characterization of Drag Coefficient of an UAV Recovery Parachute Vu Dinh Quy, Le Thi Tuyet Nhung*, Tran Minh Duy Dat Hanoi University of Science and Technology, Hanoi, Vietnam *Email: nhung.lethituyet@hust.edu.vn Abstract Parachute recovery systems are proved to be an efficient method to recovery and rescue unmanned aerial vehicles (UAV) as it follows most requirements of reliability and airworthiness in flights. Parachutes are key components of the recovery systems and the drag coefficient of parachutes plays a crucial role in evaluating parachute’s performance. The purpose of the research is to determine and compare the impact of some factors on aerodynamic drag force during the inflation of a parachute. The canopy’s shape (flat circular type and extended skirt 10% flat type), of the length of suspension lines (be in proportion to nominal diameter from 0.6 to 1.5) are considered. Measurement of the drag force of the parachute models is carried out in an open return wind tunnel. Experimental results show that flat circular canopy has a higher drag coefficient than extended skirt 10% flat model in the range of low speed from 3 to 6 m/s. However, when wind speed is greater than 6 m/s, the drag coefficients of both two parachute types are nearly 0.85. In terms of the suspension line, the longer length would significantly raise the coefficient of drag force. Keywords: UAV, parachute recovery system, parachute, drag coefficient. 1. Introduction UAV to solve most problems mentioned. Recover systems are classified based on a variety of An1 unmanned aerial vehicle (UAV) known as a constructive solutions: manual and automatic drone, is a pilotless aircraft, which is flown without command, standard or emergency procedure, shock or pilot-in-command on-board and is either remotely and without shock (mechanical, pyrotechnical/pneumatic); fully controlled from another place (ground, another maneuverable or non-maneuverable, mounted on aircraft, space) or programmed and fully autonomous. aerial systems (parachute, airbag) or ground systems UAVs initially are developed for military (net, skyhook) [2, 4]. reconnaissance, however, they are expanding their applications in other fields such as communities, law The parachute recovery system is one of the most enforcement, scientific projects, agriculture and so on popular recovery systems used for drones because its [1]. An unmanned aircraft system commonly include a benefits in operation has been proven. This system can number of cutting-edge and expensive electronic parts be designed to become a modulus equipment, which and components like sensors, central controllers, allows it to be installed directly in many kinds of batteries, actuators, optical equipment, etc [2]. In drones with a large range of take-off weight. The addition, the collected data from the flight of drones is rescue and recovery process using parachutes can be an important object which is often needed to prevent it deployed rapidly on aerial vectors so it occurs at from any unexpected access and damage. Safe and unready positions. Methods using recovery nets or airworthy requirements for components as well as skyhooks need other mechanical components and a UAVs as a whole need to take steps to maintain support group so it takes crews a lot of time to set up stability and advantages. and deploy. Besides, the parachute recovery system can slow down the speed of UAVs to the allowable Civil drones have a dramatic increase in the value while net or skyhook systems catch the drones at whole world, which leads to a challenge for authorities high speed which leads to mechanical damages of to control drones’ operations. Therefore, these could wings or propulsions. Therefore, parachutes are also have potential risks for civilians if drones are used in safer and smoother than others. However, strong wind residential or urban areas. In the future, governments is an environmental condition having effects on the can impose new regulations to guarantee that the flight parachute’s inflation and descent and hence it is more of drones would have absolutely no harm [3]. difficulty for crews to recover drones when appearing Consequently, recovery systems are gradually in complex situations [2, 5]. recognized as the most indispensable component for a ISSN 2734-9373 Received: January 5, 2021; accepted: August 25, 2021 67
2. JST: Smart Systems and Devices Volume 31, Issue 2, September 2021, 067-074 (1) (2) (3) (4) Fig. 1. The deployment process of a parachute recovery system [5]: (1) Launching a pilot chute; (2) Deploying the main chute container; (3) Opening the main chute; (4) Steady descent of UAV. Primary components of the parachute recovery factors on the parachutes’ performance. Those factors system can contain a parachute, a container for consisting of design, shape, porosity and etc. were package and launching and a control system including evaluated by experiments of various types of a receiver, a transmitter and a central processor. The parachutes with the canopy area of 1600 in2 and the whole process from starting the deployment of the same suspension line length [7]. Similarly, a program parachute to the full inflation (shown in Fig. 1) is where various types of parachute models experienced divided into four steps following: 1. Opening container drop tests was conducted at Goodyear Airdock in and launching the pilot chute; 2. Deploying the main Ohio, USA. Models which have a total canopy area of chute; 3. The main chute is fully inflated; 4. The drone 16,000 in2 (10.32 m2) were tested to compare their is in steady descent [5]. curve of the drag coefficient versus the vertical descent velocity. The report illustrated the impact caused by The drag coefficient which is dimensionless and the length of suspension lines on the drag coefficient nonlinear is a typical range of values for a specific and gliding and oscillating tendency in the motion of parachute. In steady descent, the velocity of parachutes parachutes [8]. Johari and Levshin utilized a water is constant which means that the dynamic pressure 2 tunnel to study the interaction of vortex around solid q= 1/ 2 pv is a fixed value. Hence, it can be clearly round parachute canopies. Canopy models with three seen that drag force relies on the drag area CS. For different diameters attached to a forebody were tested Do0 a given drag area, the higher value of the drag to observe the vortex in the inflation process of parachutes and measure the time-averaged drag force. coefficient C is, the smaller the nominal area S can D0 o [9] Jin et al. used the Stereoscopic Particle Image be. Thus, the drag coefficient indicates how effectively Velocimetry (Stereo-PIV) technique to investigate the a parachute canopy produces drag with a minimum of flow field around many different parachute canopies. cloth area thereby minimizing weight and volume. The The results illustrated that the geometry of parachutes drag coefficient is dependent on many nonlinear had a considerable impact on the flow structure of the characteristics so designers need to consider carefully turbulent wake. Meanwhile, the Reynold number was to give a drawing with a good balance of factors [6]. pointed out that it did not affect the profile of velocity, vorticity and Reynolds number [10-12]. In many decades, there is a large amount of research investigating all characteristics of parachutes Moreover, numerical solutions have rapidly and the flow field around them by simulation and developed and been applied to research and verify experiment manners. Air Materiel Command aerodynamic and design characteristics of parachutes. performed a series of drop and 12-feet vertical wind Stein et al. applied a parallel computational method to tunnel tests to research the effect resulting from many simulate 3-D parachute fluid-structure interaction for a 68
3. JST: Smart Systems and Devices Volume 31, Issue 2, September 2021, 067-074 round parachute (T-10 personal parachute). In detail, a deforming-spatial-domain/ stabilized space-time (DSD/ SST) finite element formulation was used for fluid dynamics, while structure dynamics was solved by a finite element formation obtained from the principle of virtual work. Because of the incompatibility of FD and SD meshes, the coupling problem was dealt with by an advanced algorithm [13]. Yu Li et al. developed the simplified arbitrary Lagrangian-Eulerian fluid-structure interaction (SALE/FSI) method to simulate the inflation of a parachute which was built as a star-shaped folded model. The research calculated and showed the number of numerical results such as opening load, drag performance, swinging angle, etc. which are in accordance with those from wind tunnel tests. Moreover, the method predicted the time-depended change in canopy shape, motion, structure and flow field around the parachute [11]. The arbitrary Lagrangian-Eulerian (ALE) approach continues to be Fig. 2. Forces acting on a parachute in the steady used and combined with interface tracking methods to descent simulate the supersonic parachute inflation in the work In steady descent, the unaccelerated parachute is of Xue Yang et al. Results obtained from numerical an equilibrium of the forces, which means that the total models such as maximum Root Mean Square (RMS) weight is equal to the drag of parachutes. Thus, the drag, general canopy shape and the smallest canopy canopy surface area S can be calculated [6]: projected areas in the terminal descent state were 0 consistent with experiments carried out in the wind 2W = T tunnel [14]. S0 2 (3) ρCvDo In this paper, measurements of the drag force of 2 parachute models with the two different types of In case, we consider the area SD0 = π o /4. Thus, canopies and the change in the suspension line length the equation to determine the nominal diameter is: will be carried out in a wind tunnel. The results are used to calculate the drag coefficient and characterize 8W D = T (4) the effects of some design factors on the drag o πρCv2 coefficient. Do 2. Theory 2.2. Drag Coefficient 2.1. Calculating the Size of Parachutes The drag force is the most important characteristic to evaluate the parachute’s performance. Let consider a parachute system in descent, that As mentioned in previous sections, a high drag is applied by the total drag of the parachute, the load, coefficient can optimize the weight and the volume of DT , and the weight of the load and the parachute parachutes. It is vital that an additional system is installed on the drones. Depending on the object using assembly, WT [6]. the parachute, the consideration to choose the type of       canopy would be discussed. The circular, flat, solid WTT+= D ma. (1) textile parachutes give superior performance The total drag force is contributed by both characteristics in drag at the descent with low speed. payload and parachute. The drag of load Dl can be While slotted parachutes could be mainly used for the neglected in relation to the large drag of parachute D application in the supersonic rate of descent, others p have gliding mode which is in accordance with which could be calculated by the equation: amphibious tasks in both military and civil 1 applications. D= ρ vC2 . S (2) p 2 Doo Previous tests indicate that a decrease of canopy porosity and an increase of suspension-line length L where ρ is air density; vC, and S are the velocity, e Do o are the prime reasons causing the growth of inflated nominal drag coefficient and area of parachutes, canopy diameter and are associated with rising drag respectively [6]. coefficient. Normally, if these lines increase their length, the canopy opens wider with the larger inflated 69
4. JST: Smart Systems and Devices Volume 31, Issue 2, September 2021, 067-074 area S p and projected diameter Dp . Drag coefficient category of T. W. Knacke [6]. All of the models’ 2 can be added if the ratio of suspension-line length and canopies had the same surface area, 0.042 m so their nominal diameter was equal to each other. The number nominal diameter D is up to 2.0. Nevertheless, in o of gores and the suspension lines was 16. The lines practice, there is a given optimized length of were attached to a stationary mounting system suspension line for a specific parachute, which means including a RC benchmark loadcell Model 1520. The that there is only a slight increase in drag force when sensor allows measuring a force up to 5 kg. extending the suspension lines [6]. 2.3. Experiments of Parachutes in the Wind Tunnel In practice, some experiments are conducted to figure out the aerodynamic characteristics by full-scale parachute models. The most popular manner is drop tests in which prototypes are deployed and inflated in the natural environment. Besides, experiments in wind tunnels are often performed due to their outstanding benefits. Wind tunnel tests allow obtaining data about the performance of parachutes with controlled environmental conditions and supporting equipment Fig. 4. Flat circular (left) and extended skirt 10% which is easier to operate and monitor than flight tests (right) canopies (top view) [15]. Therefore, this method is an effective application to compare different types of parachutes by measuring In the first series of experiments for evaluating the coefficient of force (lift, drag, tangential, ), angle the influences of the canopy shape on the drag of oscillation and so on [6]. coefficient, the canopies of parachute models were constructed by the same fabric, MIL-C-7030 Type I known as ripstop nylon 1.1 oz. The suspension lines were constructed by nylon strings with the diameter of 1mm and the length of 0.23m, which resulted in a ratio between the length of suspension lines and the nominal diameter equal to 1( LDeo/ ). The aim of the second series is to study the relationship between the length of suspension lines and the drag coefficient. Each model of two canopy types in this series maintained most specifications of the canopy in the previous series, there was no change in materials applied for the components of the model. Except for the suspension line length, the ratio Fig. 3. A parachute in flight-qualification testing in the between it and the nominal diameter was wind tunnel at NASA Ames Research Center, Calif. LDeo/ (2009) Source: NASA/JPL-Caltech. changed in the range from 0.6 to 1.5, the drag force of each model was measured with a wind speed of 8 m/s. Depending on primary purposes of tests, a kind of wind tunnel is selected to experience parachute 3.2. Experimental Setup models. The wind tunnel which is suited for collecting In the present study, the parachutes were tested in qualitative aerodynamic data with high accuracy is a low-speed wind tunnel having similar specifications closed-throat (or closed test section), full-return due to as the wind tunnel used in the study of Jin et al. at a uniform velocity distribution in the test section. Tongji University [10,12] (Table 1). However, in the present research, due to lack of facilities, the open-return wind tunnel with a closed Table 1. The specification of wind tunnel optically transparent test section is utilized. Therefore, generally, a drawback appearing in the experimental Dimensions (m) 1.0 × 0.4 × 0.5 process is not convenient for changing the parachute Maximum wind speed ≈ 0.1M (30 m/s) configuration and models [6]. 3. Experimental Setup The experimental setup of parachute testing is 3.1. Parachute Models presented in Fig. 5 while loadcell was connected to a data-acquisition circuit and then a personal computer The investigations were performed with two installed data-acquisition software. Before starting solid cloth canopy parachutes namely a flat circular experiments, wind tunnel and loadcell need to be and an extended skirt 10% (Fig. 4) based on the calibrated. 70
5. JST: Smart Systems and Devices Volume 31, Issue 2, September 2021, 067-074 In the first part, each model of two parachute types, flat circular canopy and extended skirt 10% flat canopy, was tested with the freestream velocity changing in a range of the 3 m/s to 12 m/s. At a certain speed, about 5000 samples of drag force which the loadcell transfers to the computer as the pulses of the signal were taken to calculate time-average drag force and finally drag coefficient was derived from the experimental force by equation (2). In the second part, all models with the ratio LDeo/ between a 0.6 and 1.5 are put into the uniform flow field with a speed of 8 m/s. The method to collect data about the drag force and determine the coefficient is similar to the first part. The average value of drag coefficients for each ratio between the length of the suspension line and the nominal diameter was compared to the figure when LDeo/ was equal to 1.0. 3.3. Validation The flat circular parachute is as the model which validated the accuracy and compatibility of the experimental setup for determining the drag force of parachutes in horizontal wind tunnels. In the case of determining only the drag coefficient of the flat circular parachute with the rate Fig. 5. Experiment setup of parachute testing in the of LD/1= (Fig. 6), experimental results generally wind tunnel. eo show a decreasing tendency when the freestream velocity increases from 3 m/s to 12 m/s. The coefficients of drag force corresponding to the wind speed over 4 m/s are approximate to the results presented in [7, 8] with percent errors less than 10%. For the values of drag coefficient in the experiments to consider the influence of the suspension-line length, Figure 7 shows that the change of the drag coefficient measured in the present study is the same as the results published in [8]. errors remain under 10%. Therefore, the method and setup of this research can be suited for studying further aerodynamic characteristics of parachutes. Fig. 6. The drag coefficient vs. velocity plot of the flat circular parachute Fig. 8. The drag coefficient vs. velocity plot of two Fig 7. The drag coefficient vs. the rate of LDeo/ plot types of canopies. of the flat circular parachute. 71
6. JST: Smart Systems and Devices Volume 31, Issue 2, September 2021, 067-074 4. Results and Discussion At the freestream velocity lower than 5 m/s, there is a considerable difference in the drag coefficient of 4.1. Influence of Canopies the two canopy types. To be more specific, the figure The condition of the atmosphere maintained at of the flat circular canopy, which is 1.72, is more 0.27 the temperature of 25± 1 oC and pressure of than of the extended skirt 10% flat canopy. However, 1.01× 105 Pa. Figure 8 displays the comparison of the at the high velocity, the extended skirt 10% flat canopy curve of the drag coefficient versus the freestream has higher time-averaged drag forces than the flat velocity between two types of parachute canopies. circular canopy although the different level in drag There is a decreasing trend in the drag coefficient if the coefficients of two parachute types is pretty small velocity increases, the drag coefficient gradually went around 0.1. down to about 0.85 at the inlet stream over 10m/s. At the low speed, the reason for the canopies’ instability is that pressure on the inner canopies’ surface is not uniform and high enough to maintain the stable structure of canopies and the high tension of the suspension lines. Hence, the skirt of canopies can be expanded, that is the projected area S p is larger than the nominal area So and a little drag force is added to the total one. In contrast, high pressure putting on the inner surface at high speeds makes the canopies more stable, the projected area S p is approximate to the nominal area So . Therefore, the drag coefficient at the low inlet velocity is greater than at the high values with the constant nominal area. Fig. 9. Flat circular canopy at various freestream The difference in the drag coefficient of two velocities, a) 3m/s, b) 6m/s, c) 9m/s, d) 12m/s. canopy types is derived from the stability of the canopy structure when the turbulence occurs. If we considered the parachutes and the flow field in the test section along the horizontal axis, there is the turbulent flow occurring behind the canopies. In terms of the flat canopy, the maximum projected diameter is the diameter at the edge of the canopy skirt. The turbulent flow affects directly that edge, which makes the canopy expand. Meanwhile, the flat canopy’s edge with the extended skirt 10% is shrunk, the diameter at the edge is smaller than the maximum inflated diameter of the parachute. Thus, this protects the parachute’s structure from the negative effect of the turbulence, the amount of inner pressure is controlled and not changed significantly. Fig. 10. Extended skirt 10% flat canopy at various freestream velocities, a) 3m/s, b) 6m/s, c) 9m/s, d) In addition, when the wind speed was larger than 12m/s. 8 m/s, the oscillation of parachutes appeared and became more intensive. Two parachute models also oscillated with small angles at a speed less than 6 m/s. 4.2. Effect of the Length of Suspension Line The experimental process was performed in the air at 22± 2 oC and 1.01 ×1 05 Pa. Since the freestream velocity is kept at the constant medium value of 8 m/s, the oscillation of parachutes is not too large. Figure 11 illustrates the coefficient of drag force grows if the length of the suspension lines increases, which means there is an increase in the ratio between that length and the nominal diameter. For the range of the ratio LDeo/ between 0.6 and 1.0, the drag coefficient of two types goes up significantly, the additional proportion of the Fig. 11. The percent change in drag coefficients of two flat circular canopy is less than of the extended skirt types of parachute canopies 72
7. JST: Smart Systems and Devices Volume 31, Issue 2, September 2021, 067-074 10% flat canopy. shape and the suspension-line length on the drag coefficient of parachutes were investigated. The Nevertheless, it is witnessed that there is a results show that the flat circular canopy produced the remarkable difference in the plot of the percent change drag force little less than the extended skirt 10% of the drag coefficient between two models when models when the freestream velocity is over 5 m/s. LD/ is from 1.0 to 1.5. While the figure of extended eo Based on the observation during the experimental skirt 10% flat models only increases slightly and process, it is concluded that the parachutes applying in reached nearly 4.4%, the flat circular canopy recovery systems in UAVs should be designed to have experienced a dramatic growth in the drag coefficient. the rate of descent from 5 - 7 m/s. That rate guarantees At LDeo/= 1.5 , the increasing percentage of the flat the safety requirement for UAVs and the performance circular parachute in comparison with the ratio 1.0 was of parachutes, that is the small weight, high drag and about twice as much as the proportion of the extended stability with small oscillation. skirt one. With 0.6<<LDeo / 1.0 , two types of parachutes In previous studies, Knacke proved that the drag have the increasing trend in percent change of drag coefficient of the flat circular parachute was possible although the figure of the flat circular canopy is to increase continuously for LDeo/ up to 2.0 [6]. smaller than the extended skirt 10% parachute. Meanwhile, the extended skirt 10% flat has little However, in this case 1.0<<LDeo / 1.5 , the growth in its drag coefficient, approximate to 2% at the additional percentage of drag coefficients of the flat ratio LDeo/ greater than 1.3 [6], [16]. It is difficult to circular model is more than the extended skirt 10% flat determine exactly small developments in drag force parachute. due to the limit of the accuracy in the measurement by The paper only investigated some factors which loadcell, in addition, the flexibility of canopy and have the impact on the drag coefficient of parachutes. instability of flow field through it have a significant It is necessary to investigate other factors to consider impact on the drag force. Therefore, there are some abnormal points in the curve such as the figure at completely the characteristics and performance of parachutes before applying them for recovery systems LDeo/ = 1.4 on the plot of the flat circular canopy and in UAVs. Furthermore, future studies can take into at 1.5 on the curve of the extended skirt 10% consideration the porosity of material and stability in parachute. the descent motion, the parachute models are The change in the line length causes the recommended such as cross, ribbon or disk-gap band expansion or the shrinkage of the parachute canopies. parachutes. As mentioned above, the specific kind of parachute has the optimizing length for its suspension lines, whose References ratio LDeo/ to the nominal area is about 1.0. [1]. Unmanned Aircraft Systems (UAS), International Declining the suspension lines’ length leads to Civil Aviation Organization, Montréal, Canada, 2011, shrinking the area of canopies, as a result, the drag pp. 3. force reduces certainly. In addition, the forebody wake created by the mounting stationary also contributes to [2]. Marcus Eriksson, Patrick Ringman, Launch and decreasing the total drag force. recovery systems for unmanned vehicles onboard ships, Centre for Naval Architecture, 2013. The primary reason to explain the difference [3]. Operation of Small Unmanned Aircraft Systems Over between two curves of the parachute in Fig. 11 is People, Federal Aviation Administration, Department similar to the previous cases (4.1. Influence of canopy Of Transportation, USA, 2019. shapes). If the suspension lines are extended, consequently the canopies will inflate wider and [4]. T. Wyllie, Parachute recovery for UAV systems, Aircraft Engineering and Aerospace Technology, vol. simultaneously there is a dramatic growth in the drag 73, pp. 542-551, 2001. force. The inflation of the flat circular canopy totally follows the phenomenon. However, the structure of the extended skirt 10% flat canopy only allows an [5]. Vasile PRISACARIU, Sebastian POP, Ionică expansion to a certain dimension. To be more specific, CÎRCIU, Recovery system of the multi-helicopter uav, the lower part or the skirt whose diameter is equal to in Review of the Air Force Academy, Braşov, 0.1D restricts the opening of the canopy. Therefore, Romanian, Henri Coanda Air Force Academy o Publishing House, 2016, pp. 91-98. the projected area of that canopy is nearly fixed, in other words, the amount of the produced drag force remains stable or increases slightly. [6]. T. W. Knacke, Parachute Recovery Systems, China Lake, CA, US, 1991. 5. Conclusion [7]. T. W Knacke and A. M. Hegele, Comparison Tests of In conclusion, in this paper, the effects of canopy Various Types, U.S. Air Force,USA, Tech.Rep. Model 73
8. JST: Smart Systems and Devices Volume 31, Issue 2, September 2021, 067-074 Parachutes, 1949. Simulation Techniques, Journal of Aircraft, p. 802, 2001. [8]. J. F. Stimmler and R. J. Ross, Drop Tests of 16,000 sq. in. Model Parachutes, U.S. Air Force, USA, Tech.Rep. Model Parachutes, Volume VIII, 1954. [14]. Xue YANG, Li YU, Min LIU, Haofei PANG, Fluid structure interaction simulation of supersonic [9]. H. Johari, A. Levshin, Interaction of a line vortex with parachute inflation by an interface tracking method, a round parachute canopy, Journal of Fluids and Chinese Journal of Aeronautics, vol. 33, pp. 1692- Structures, p. 1258–1271, 2009. 1702, 2020. [10]. Zhe-Yan Jin, Sylvio Pasqualini, Bo Qin, Experimental [15]. DJ.Cockrcll, The Aerodynamics of Parachutes, List of investigation of the effect of Reynolds number on flow National Distribution Centres, NATO Science and structures in the wake of a circular parachute canopy, Technology Organization. , 1987. Acta Mechanica Sinica, 2014. [16]. T. W. Knacke, L. L. Dimmick. , Design Analysis of Final Recovery Parachutes for B- 70 Encapsulated [11]. Yu Li, Cheng Han, Zhan Ya’nan, Li Shaoteng, Study Seat and the USD-5 Drone, Ohio : USAF, 1962. of parachute inflation process using fluid–structure, Chinese Journal of Aeronautics, vol. 27, pp. 272-279, [17]. Ashim Panta, Simon Watkins, Reece Clothier, 2014. Dynamics of a small unmanned aircraft parachute system, Journal of Aerospace Technology and Management, vol. 10, 2018. [12]. Zheyan Jina, Sylvio Pasqualini, Zhigang Yang, Experimental investigation of the flow structures in the wake of a, European Journal of Mechanics B/Fluids, [18]. Sylvio Pasqualini, Zheyan Jin, Zhigang Yang, pp. 70-81, 2016. Measurement of the flow structures in the wakes of different types of parachute canopies, Acta Mechanica Sinica, 2017. [13]. Keith R. Stein, Richard J. Benney, Fluid-Structure Interactions of a Round Parachute: Modeling and 74