thrust measurement of r/c scale jet engine using cfd and hardware testing

Thrust Measurement of R/C Scale Jet Engine Using

CFD and Hardware Testing

Je Won Hong, Ovini Silva, Ankur Mehta,∗

Kevin Burns, William Yoshida, David Liljewall, Michael Vershaw,∗

Indronil Ghosh,†

and Bradley Rafferty‡

University of Illinois at Urbana-Champaign, Urbana, IL, 61801

A turbojet engine is a complex and precisely engineered device that relies on multiplestages of machine-fluid interaction before ultimately producing the desired output thrust.The engine’s internal hardware - fan, compressor, combustor, turbine, and mixer - wouldnever achieve this propulsive force if not for a nozzle that dictates the thrust level. This re-search paper details the efforts and findings of the AIAA JetCat Engine Technical ProjectGroup of University of Illinois at Urbana-Champaign in its endeavor to analyze the effectsof nozzle design on the thrust performance of model jet engines through computationalfluid dynamics (CFD) software. To test the validity of the CFD results, the measuredthrust is compared to manufacturer specifications. For further validation of physical per-formance, the JetCat P140-RX Miniature Turbine Engine is mounted to two pillow blockbearings riding two parallel rails to allow for unhindered, uniaxial movement and to pro-vide an accurate thrust reading via spring loading scales. The team has also designedthree nozzles for the engine on computer aided design (CAD) software to compare themagainst each other as well as against data from the stock nozzle, specifically peak thrustand various flow characteristics. The purpose of these efforts is to gain valuable experiencein research, computational fluid dynamics, properly controlled testing, and hands-on workwith a miniature jet engine.

Nomenclature

dS differential surface area elementdV differential volume elementdy differential in terms of ydz differential in terms of zm mass flow rateγ specific heat ration normal unit vectorρ density of substance~F net force due to stresses on control volume surface~f body force per unit mass~T thrust~u velocity of flow~v velocity of control volume in motionA cross-sectional areac speed of sound in airM Mach number

∗Department of Aerospace Engineering, University of Illinois at Urbana-Champaign, AIAA Student Member†Department of Physics, University of Illinois at Urbana-Champaign, AIAA Student Member‡Department of Mechanical Engineering, University of Illinois at Urbana-Champaign, AIAA Student Member

1 of 11

American Institute of Aeronautics and Astronautics

P pressurep pressureR specific gas constantT temperature

Subscripts

a aire exitf fuelin inlet

I. Introduction

Computational Fluid Dynamics is a useful engineering tool that allows the estimation of performanceparameters and validation of a certain device that is a subject to testing. Two dimensional flow analysis

of an exhaust nozzle is an important example of computer-based simulation which requires understandingof: physical theory and phenomena of incompressible and compressible flow, ideal gas theory, and controlof proper meshing techniques and appropriate boundary conditions. The AIAA Student Chapter at theUniversity of Illinois Urbana-Champaign focuses on the validation of its small-scale jet engine utilizing bothcomputer simulation and hardware testing using a test stand designed and built by a group of students.According to CFD results, theoretical estimation of the JetCat P140-RX engine thrust closely matches themanufacturer specification within 9% error, although the sophistication of our physical validation is not yetfinalized. Further validation is to be carried out by using an electronic force sensor installed on the enginetest stand and running the engine at different throttle levels.

II. Theoretical Calculation of Thrust

General thrust of a turbo-machine follows conservation of mass and momentum, in which the followingequations are valid. Equation (1) and Equation (2) are the conservation of mass and momentum, respec-tively.1

∂

∂t

˚

V

ρ dV +

‹

S

ρ~u · n dS = 0 (1)

∂

∂t

˚

V

ρ~u dV +

‹

S

ρ~u (~u− ~v) · n dS = −

‹

S

ρ n dS +

˚

V

ρ~f dV + ~F (2)

Assumption of steady flow, stationary control volume and constant pressure acting on the cross-sectionalarea reduces the conservation of mass and conservation of momentum equations to the following.

¨

S

ρ~u · n dy dz = 0 (3)

¨

S

ρ~u (~u− ~v) · n dy dz +

¨

S

p n dy dz = ~F (4)

Equations (3) and (4) constitute the analytical expressions by which thrust is calculated. A turbojetengine consists of an air inlet, a compressor, combustion chamber, turbine blades, and a nozzle exit as shownin Figure 1. Our simplification of the jet engine subject to testing is based on the assumption that it consistsof an inlet and outlet with a pressure drop due to the combustion process. This simplification results in ageneral thrust equation with known parameters at the inlet and outlet. Mass and momentum conservationexpress the mass flow rate at the nozzle exit and the resultant thrust as follows.

m = ρ ~uA (5)

me = ma + mf = min + mf (6)

~T = meve − minvin + (Pe − Pin)Ae (7)

2 of 11


Figure 1. Diagram of jet engine in axial cut

Figure 2. Diagram of cut-view of JetCat engine2

With known quantities specified by the man-ufacturer of the engine and by CFD results, thethrust of the device may be estimated using theabove expressions. The estimated thrust is to becompared to the manufacturer specification to val-idate our results.

The JetCat P140-RX R/C scale jet engine isknown for its simplicity of operation, its stabil-ity, and its relatively low fuel consumption rate.The simple nozzle geometry and assembly featurealso serve as a perfect test subject for 3D CADprogramming, nozzle design manipulation, and 2DCFD analysis. From the cross-sectional cut of Jet-Cat engine shown in Figure 2, it is reasonable tomake the assumption of a simplified jet engine withcomponentry as shown in Figure 1.

III. CFDSimulation Using ANSYS Fluent

A. Nozzle Measurement and GeometrySet-Up for CFD

ANSYS Fluent 15.0 is the CFD simulation tool ofchoice due to its simple user interface and power-ful meshing ability in 2D flow analysis.3 ANSYSFluent is capable of importing the geometry of asimulation subject into its Workbench Modeler andaccepting the subject as a surface in which a fluidof choice can be physically simulated by assigningboundary conditions for inlet type, outlet type, and a wall section. To properly design the nozzle geometry,the engine nozzle is disassembled and its dimensions are measured. Based on dimensions tabulated in Ta-ble 1, the nozzle is modeled using Creo Parametric 3.0 and its cross-sectional geometry is imported to ANSYSWorkbench. The nozzle geometry in 3D CAD model format is shown in Figure 3, and its cross-section thatwas implemented in ANSYS Workbench is shown in Figure 4.

Figure 3. JetCat P140-RX nozzle in 3D CAD modelformat

Figure 4. Cross-sectional nozzle sketch

3 of 11


B. ANSYS Modeler and Mesh Set-Up4

Table 1. JetCat P140-RX Nozzle Dimen-sions

Dimension Type Measurement (mm)

Inlet Diameter 36.0

Length 56.0

2D nozzle flow simulation on ANSYS is set up as a nozzle-to-chamber configuration in order to observe the flow characteris-tics past the exit area of the nozzle. The chamber dimensionsare 47.5X100 mm, as shown in Figure 5. The surface createdfrom the sketch is divided into 14 different sections to allow moreprecise meshing at the nozzle wall and chamber wall boundaries.The nozzle has the rising point at the coordinate of 26.0 mmfrom y-axis with offset of 7.5 mm, combined with a curvature that connects two lines drawn toward the inletand outlet respectively. Meshing the simulation subject properly is one of the most critical stages in anyCFD analysis. Improper meshing can cause the simulation result to not converge properly and significantdiscrepancies will arise when compared to the accepted values, e.g. manufacturer specifications. For thenozzle simulation, the mesh size is set to .25 mm for all sections from the model sketch. It is important toproperly assign the inlet, outlet, wall, and axial boundaries. Below in Figure 6 is the resultant mesh.

Figure 5. ANSYS Workbench geometry implementationof JetCat nozzle and chamber

Figure 6. Meshing of the nozzle and chamber

C. Boundary Conditions and Simulation Model

According to the manufacturer specifications (see Table 5 of the Appendix), the JetCat P140-RX has apressure ratio of 3.2 at its maximum throttle. For our purposes, we assume that the pressure inside thetesting chamber is equal to atmospheric pressure and its temperature is room temperature. The specificationsalso state that the maximum temperature of the gas flow from JetCat nozzle reaches 950 K. Ideal gas andlaminar flow with energy equation are utilized for this particular 2D flow analysis. The axis-symmetricoption is checked to mirror the simulation result as a full nozzle and chamber combined. Below in Table 2are the boundary conditions used for this simulation.

Table 2. Boundary conditions for 2D flow simulation

Variables

Component Boundary Type Temperature (K) Pressure (kPa)

Nozzle Inlet Pressure-Inlet 950 324.24

Nozzle Outlet Pressure-Outlet 300 101.33

An example of the inlet boundary condition set-up interface is shown in Figure 7. The number of iterationsis set to 250 and the meshing technique is also iterated and modified accordingly to ensure reasonablesimulation results. The calculation process on ANSYS Fluent interface is shown in Figure 8.

4 of 11


Figure 7. Boundary condition set-up interface Figure 8. Iteration and calculation interface

D. Result and Thrust Calculation

The result of 2D flow analysis for the JetCat engine nozzle is exported in the form of a contour plot. Thevelocity, pressure, and temperature contour plots are shown in Figures 9, 10, and 11, respectively.

Figure 9. Flow velocity [m/s] contour of the JetCat nozzle and chamber

Figure 10. Flow pressure [Pa] contour of the JetCat nozzle and chamber

5 of 11


Figure 11. Flow temperature [K] contour of the JetCat nozzle and chamber

Flow velocity, pressure, and temperature on the nozzle axis are plotted as shown in Figures 12, 13,and 14, respectively. The three quantities are plotted on the domain of the axial coordinate of the nozzleto represent inlet, internal, and exit flow characteristics. The simulated parameters at the nozzle exit areused for theoretical calculation of the nozzle thrust in the next section using equations previously stated inSection II of this report.

Figure 12. Gas flow velocity along JetCat nozzle axis

6 of 11


Figure 13. Gas flow pressure along JetCat nozzle axis

Figure 14. Gas flow temperature along JetCat nozzle axis

IV. Validation of Simulation and Theoretical Calculation

Based on the assumptions of steady, one-dimensional, isentropic flow of an ideal gas, the following ex-pressions of pressure ratio and temperature ratio are derived and utilized.5

Pin

Pe=

(

1 +γ − 1

2M2

)γ/(γ−1)

(8)

Tin

Te= 1 +

γ − 1

2M2 (9)

c =√

γRaT (10)

The simulations results of ANSYS Fluent for pressure, gas flow velocity, and temperature at the nozzle’sinlet and outlet are tabulated below.

7 of 11


Table 3. Various results of flow simulation through nozzle

Quantity Type

Nozzle Part Gas Flow Velocity (m/s) Temperature (K) Pressure (kPa)

Inlet 145.95 940.63 309.69

Outlet 400.20 876.50 217.72

The Mach number of the gas flow at the nozzle exit is 0.6745, and γ is 1.369 as per Equation (9).Calculating the theoretical gas pressure at the outlet yields 290.57 kPa, which presents an error of 6.2% incomparison to the results of the simulation. The error is likely due to the simplification of the engine modelas experiencing constant pressure along its inlet/outlet cross-section, which is not true for a viscous flowmodel. This error is quite reasonable and small enough to validate the theoretical calculations.

The inlet quantities of the engine are calculated by referencing manufacturer specifications, which statethat the JetCat P140-RX has an air mass flow rate of 0.35 kg/s at the intake. The diameter of the intakeis 112 mm, from which the area is calculated as 0.00985 m2. In using Equation (5) and taking air densityat room temperature (300 K) as 1.184 kg/m3, the inlet flow velocity is calculated to be 30 m/s. By adding0.0073 kg/s of fuel consumption rate at maximum throttle of the engine, the thrust is calculated to be 130.2N. JetCat P140-RX has a maximum thrust of 142 N according to specifications, which is a mere 8.3% error.6

The engine specifications list the maximum gas exhaust velocity as 1461 km/h, or 405.83 m/s. Thesimulated result of 400.20 m/s is only 1.39% different. Therefore, our ANSYS Fluent CFD results are validand quite accurate for this simple converging nozzle performance analysis.

V. Physical Testing

The customized test stand is constructed using two linear rails and sliding pillow block bearings withminimal friction in attempt to create unhindered uniaxial motion for the physical testing of the engine.Mounting brackets, which are sold separately by JetCat, are installed on the engine body and the engine isattached to the bearings via two laser-cut, wooden adapter plates as shown in Figure 15. To halt the motionof the engine-bearing assembly at the end of the rails, two blocks of square aluminum rod are cut and boltedto the test stand at the end of the rails as shown in Figure 16.

Figure 15. Pillow block bearings, rails, and JetCat P140-RX assembly on the test stand

8 of 11


Figure 16. Stoppers installed at end of the test rails Figure 17. Assembly of JetCat engine test stand

The initial plan was to employ an electronic force sensor, however spring load scales were substituted forthis purpose due to hardware issues. Two manual spring load scales are connected to a metal spacer installedon each pillow block and are rigidly attached to the stand. This thrust measurement technique is not asaccurate as the electronic alternative, and so slight discrepancies arise when compared to the simulationresults and the manufacturer specifications. Testing is conducted in 10-minute intervals and consists ofignition, low throttle, max throttle, idle, and finally a cooling process. A total of three runs are conductedwith thrust and temperature readings recorded. Temperature is measured via a temperature sensor that isinstalled by the manufacturer within the nozzle.

Table 4. Physical test results of the JetCat P140-RX

Trial Number Temperature (◦C) Thrust on Scale 1 (N) Thrust on Scale 2 (N) Total Thrust (N)

1 725 69.0 67.0 136.0

2 731 70.0 69.0 139.0

3 727 69.0 68.0 137.0

Average 727.67 69.3 68.0 137.3

Likely sources of error are: friction between the metal rails and pillow block bearings, inexact alignmentof JetCat engine along the desired axial direction of motion, and spring load scale friction. Despite thesesources of error, the calculated error is still well within reason.

VI. Conclusion

In our efforts we were able to successfully analyze, both numerically and physically, the thrust output ofthe miniature jet engine. Our simulations were refined to the point of error within 9% or less when comparedto manufacturer specifications, and in this refinement process we learned the importance and intricacy ofiteration and convergence to a desired solution. Additionally, our physical testing confirmed the accuracyof our simulations and theoretical calculations and agrees well with expected values. The testing couldbe improved by the successful implementation of an electronic force sensor for more accurate readings, aswell as by employing a stable, one-rail testing rig to reduce the amount of friction in the system. We arenow prepared to manufacture our own nozzle and compare it against our current findings. In all, we arevery satisfied with the knowledge we have gained regarding CFD analysis, design for manufacturability, andphysical testing implementations.

VII. Future Plans

Our next step for the project is to manufacture one of the nozzles we designed. The models below inFigures 18 and 19 are examples of nozzle geometry variations we have designed that will be analyzed inANSYS, and one of which will be sent to the metal shop to be manufactured. Figure 20 shows the jetengine fitted with a 3D-printed nozzle for dimensional analysis purposes. This is the nozzle design we planto manufacture out of solid aluminum. We will also explore 3D CFD analysis in order to expand on the 2D

9 of 11


analysis we have established.

Figure 18. Nozzle design with concave-up curvature Figure 19. Nozzle design with straight line geometry

Figure 20. 3D-printed nozzle design fitted to the engine

Appendix

A. Manufacturer Specifications of JetCat R/C Scale Jet Engine3

Type P140-RX

Idle RPM (1/min) 32000

Max RPM (1/min) 125000

10 of 11


Idle thrust (N) 6

Max thrust (N) 142

EGT (◦C) 520-749

Pressure ratio 3,2

Mass flow (kg/s) 0,35

Exhaust gas velocity (km/h) 1461

Power output (thrust) (kW) 28,8

Fuel consumption at Max RPM (ml/min) 550

Fuel consumption at Idle (ml/min) 110

Fuel consumption at Idle (kg/min) 0,088

Fuel consumption at Max RPM (kg/min) 0,44

Specific fuel consumption at Max Rpm (kg/Nh) 0,186

Weight (g) 1580

Diameter (mm) 112

Length (incl. starter) (mm) 293

B. Pressure, Temperature, and Gas Flow Velocity Simulation Values from ANSYS Fluent

X [m] Pressure [Pa] Temperature [K] Velocity [m/s]

-5.00E-03 3.07E+05 9.39E+02 1.49E+02

1.78E-03 3.10E+05 9.41E+02 1.46E+02

8.56E-03 3.12E+05 9.42E+02 1.47E+02

1.53E-02 3.12E+05 9.42E+02 1.52E+02

2.21E-02 3.12E+05 9.42E+02 1.63E+02

2.89E-02 3.10E+05 9.41E+02 1.80E+02

3.57E-02 3.03E+05 9.36E+02 2.09E+02

4.24E-02 2.88E+05 9.26E+02 2.55E+02

4.92E-02 2.60E+05 9.07E+02 3.20E+02

5.60E-02 2.18E+05 4.00E+02

Acknowledgments

The authors thank Rodney L. Burton, the professor Emeritus of the University of Illinois at Urbana-Champaign Aerospace Engineering department, for advising us and reviewing the overall contents and valid-ity of test process, Greg S. Milner and Lee A. Booher of the Aerospace Engineering department machine shopfor their generosity regarding hardware components for test stand assembly and manufacturing knowledgefor the nozzle, and Kevin Kim and Brian S. Woodard of the Aerospace Engineering department for CFDsimulation instructions.

References

1Anderson, J. D., Fundamentals of Aerodynamics, McGraw-Hill, US, 5th ed., 2010.2Jacob A. Baranski, John L. Hoke, P. J. L. and Schauer, F. R., “Preliminary Characterization of Biofuels using a Small

Scale Gas Turbine Engine,” 49th AIAA Aerospace Sciences Meeting, January 2011.3GmbH, M. Z., JetCat RX Turbines with V10 ECU , Ingenieur-Bro CAT, Staufen im Breisgau, Germany.4Kishore, G. B. and Akash, K., “CFD Analysis of a Rocket Nozzle with one Inlet at Mach 0.6,” Tech. rep., Aeronautical

Engineering, Malla Reddy College of Engineering and Technology, Hyderabad, Project Report.5Seitzman, J. M., “Isentropic Flow with Area Change,” Georgia Institute of Technology, PDF, 2001.6Gerald Hagemann, Richard Schwane, P. R. and Ruf, J., “Plug Nozzles: Assessment of Prediction Methods for Flow

Features and Engine Performance,” 38th AIAA/ASME/SAE/ASEE Joint Propulsion Conference & Exhibit , 2002.

11 of 11
