Rotorcraft Engine- Nacelle Cooling Model Calibration Project Nacelle Cooling Solutions Senior Design Team Mechanical Engineering College of Engineering and Natural Sciences

19 November 2004NCS Nacelle Cooling Solutions: The Team Jason Lee Team Leader Theran Cochran Team Secretary Colby Huffmon Team Financial Officer David Tallman Productions Manager Technical Advisor: Dr. Earl P. N. Duque

19 November 2004NCS Project Sponsor Boeing Company: Integrated Defense Systems Mesa, AZ Rotorcraft Division Mr. Wesley Ussery, B.S.M.E. NAU Mr. Wesley Ussery, B.S.M.E. NAU Dr. Nathan Adams Dr. Nathan Adams

19 November 2004NCS Who is Boeing? World’s Largest Manufacturer of Military Aircraft. Sales in 2003 were a reported $50.5 billion to customers in 143 countries. 56% was in Defense Systems. © Boeing Company

19 November 2004NCS Presentation Overview Project Objectives Breakdown of tasks Discussion of Computational Model Discussion of Experimental Model Our Vision of the project’s future

19 November 2004NCS Industry Standard Model Methodology for engine cooling analysis is described in SAE, ARP-996A, “Cooling Data for Turbine Engines in Helicopters”. Originally written in 1967, and last revised in 1986.

19 November 2004NCS What is ARP-996A? Describes a standard method of presenting needed data and calculating the required cooling air for a given engine-nacelle installation in rotorcraft. “Purpose: Efficient design of a turbine engine installation requires … Cooling margins developed by these methods would be subject to full scale testing for verification.”

19 November 2004NCS Project Statement Our Objective: Determine a confidence interval to be associated with results obtained from the industry standard model. Our study is based upon the AH-64 installation of the Apache Longbow helicopter.

19 November 2004NCS AH-64 Data

19 November 2004NCS Project Execution Three Main Phases: Computational Model Development Computational Model Development Experimental Development Experimental Development Results and Recommendations. Results and Recommendations. Current Status: Completing Computational Stage and beginning Conceptual stage of the test model.

19 November 2004NCS Computation: Phase I Major Tasks Understanding the underlying theory behind the model described by ARP-996A. Understanding the underlying theory behind the model described by ARP-996A. Develop a numerical algorithm for the model. Develop a numerical algorithm for the model. Implement a computer program to execute the algorithm. Implement a computer program to execute the algorithm.

19 November 2004NCS Experimental: Phase II Major Tasks Develop an appropriately scaled model of the engine-nacelle installation. Develop an appropriately scaled model of the engine-nacelle installation. Design and execute an appropriate experiment. Design and execute an appropriate experiment. Analyze experimental data and determine a confidence interval. Analyze experimental data and determine a confidence interval.

19 November 2004NCS Results: Phase III Major Tasks Based on results of data analysis, determine a recommendation for improvements, and/or advice on interpretation of results from ARP- 996A methodology. Based on results of data analysis, determine a recommendation for improvements, and/or advice on interpretation of results from ARP- 996A methodology. i.e. a fudge factor for the methods described in ARP-996A i.e. a fudge factor for the methods described in ARP-996A

19 November 2004NCS Computational Model Used to provide numbers for comparison with experiment Based on the model described in SAE ARP-996A Engine is broken lengthwise into several elements Energy balance on each element

19 November 2004NCS 1-D Model Schematic

19 November 2004NCS Nodal Energy Balance Equations Engine surface: Nacelle: Annulus flow:

19 November 2004NCS Solving the Energy Balance for Each Element Energy balance equations Three equations Three equations Non-Linear Non-Linear Use Newton’s Method for Non-Linear Systems

19 November 2004NCS Newton's Method for Nonlinear Systems Given a vector of n functions, find simultaneous roots for all of them The messy part: calculating the Jacobian matrix

19 November 2004NCS Newton's Method for Nonlinear Systems Solve linear system (J(x))y = F(x) Gaussian elimination or Cramer's rule Gaussian elimination or Cramer's rule ARP uses Cramer's rule ARP uses Cramer's rule Easiest to just use \ operator in Matlab Easiest to just use \ operator in Matlab set new x = x + y repeat until y is close to zero

19 November 2004NCS Find T 1, or W? ARP uses mass flow rate of the annulus as one of the variables in the node equations Using the engine surface temperature instead has advantages Mass flow rate must be the same for each node Mass flow rate must be the same for each node Temperature can change Temperature can change The math is simpler The math is simpler Required mass flow rate can still be found Required mass flow rate can still be found

19 November 2004NCS Finding the Required Mass Flow: The ARP Way 1.Calculate T 2, T a, W for first element 2.Calculate T 2, T a, W for next element 3.Take maximum W 4.Re-Calculate temperatures of previous elements 5.Repeat from 2. for each element 6.Re-calculate required flows from step 1. until converged

19 November 2004NCS Finding the Required Mass Flow: The New Way 1.Make a guess for the required mass flow W 2.Calculate temperatures throughout engine 3.Are the temperatures all low enough? if yes, then the flow rate is high enough if yes, then the flow rate is high enough if no, then increase the flow rate and try again if no, then increase the flow rate and try again

19 November 2004NCS Advantages of the New Way Flow rate is automatically held constant over the entire engine Easier to non-dimensionalize the node equations Easier to calculate the Jacobian matrix Don’t have to deal with changing h with W Don’t have to deal with changing h with W

19 November 2004NCS Non-dimensional Nodal Energy Balance

19 November 2004NCS Test Model Development Based on data for the AH-64 installation, a simplified model can be described. A series of cylinders, with nominal diameters given by scaled AH-64 data.

19 November 2004NCS Our Physical Model Geometry

19 November 2004NCS Physical Model Concepts Scale 1:2, 6061 Aluminum to be used, or 15 gauge sheet metal Nacelle circular cross-section to simplify airflow velocity profiles

19 November 2004NCS Experimental Heat Source Concepts Resistance wire and a current source. Propane burners

19 November 2004NCS Engine with Nacelle

19 November 2004NCS The Next Steps to Our Goal Material Selection Heating Element Selection Model Construction Test Rig Design and Construction Data Acquisition Execution

19 November 2004NCS Phase II: Schedule Phase II Experimental Development: Including Test model development Design of Experiment Procurement and Construction Experiment Execution: Including Data Analysis

19 November 2004NCS Experimental Development What are we trying to achieve? What will the measurements be? What will the measurements be? Engine Surface Temperature Nacelle Surface Temperature Cooling Air Temperature How will we get the data from the experiment? How will we get the data from the experiment? Appropriate Data Acquisition

19 November 2004NCS Data Analysis What do we do with the data when we’ve run the experiment? Compare surface temperature profiles with those obtained from the computational model. Compare surface temperature profiles with those obtained from the computational model. Based on this comparison, determine the confidence interval for the methods described in ARP-996A. Based on this comparison, determine the confidence interval for the methods described in ARP-996A.

19 November 2004NCS Results and Recommendations Based upon the results from the data analysis, we can recommend one of two things: A revision to ARP-996A, consisting of the addition of a warning section describing the accuracy of the methods described there in. A revision to ARP-996A, consisting of the addition of a warning section describing the accuracy of the methods described there in. A complete revision of ARP-996A, including a new model describing new methods. A complete revision of ARP-996A, including a new model describing new methods.

19 November 2004NCS In Conclusion: What are we trying to accomplish? A measure of “goodness” for the 1-D model described in SAE, ARP-996A. A measure of “goodness” for the 1-D model described in SAE, ARP-996A. Provide data from an appropriate experimental test to back up our conclusions. Provide data from an appropriate experimental test to back up our conclusions.

19 November 2004NCS Any Questions?

19 November 2004NCS What is a Nacelle???? A streamlined enclosure for an aero-engine installation.

19 November 2004NCS AH-64 Apache Longbow Helicopter

19 November 2004NCS

19 November 2004NCS

19 November 2004NCS

19 November 2004NCS Propulsion Air Management System Review the AH-64 installation as a typical example of a helicopter propulsion system. AH-64 data will be the basis for experimental study and validation.

19 November 2004NCS The AH-64 Installation

19 November 2004NCS Computational Model Create computer code to solve 1-D model described in ARP-996A. Provide numerical data for comparison with that obtained from experiment.

19 November 2004NCS 1-D Description

19 November 2004NCS Experimental Model Develop a physical model of the AH-64 Installation to be tested Experimental data will be used to validate and provide “goodness” of the computational model.

19 November 2004NCS Preliminary Project Schedule