NASA Technical Reports Server (NTRS) 20150002943: Development of an Aerodynamic Method and Database for the SLS Service Module Panel Jettison Event Utilizing Inviscid CFD and MATLAB

2.5 MB·English

by  NASA Technical Reports Server (NTRS)

Development of an
Aerodynamic Method and
Database for the SLS Service
Module Panel Jettison Event
Utilizing Inviscid CFD and
MATLAB
Mike Applebaum/CRM
Les Hall/CRM
Marc Eppard/CRM
Dave Purinton/Gray
John Blevins/MSFC
Ricky Campbell/MSFC
www.nasa.gov
Introduction
Goal: Create an analysis tool which can be coupled with a 6-DOF model to accurately
predict SM panel separation from SLS in a time efficient manner.
•Requires spatial prediction of SM panel flight space in proper environment,
generated with CFD analysis. The resultant database is divided into three zones:
• panels on the hinge during initial separation in which not only body, but panel
to panel effects are important
• panels in near proximity to the body
• panels alone in freestream environment
• Data placed into Matlab, which utilizes the interpolation routines
• Coupled with 6-DOF, which includes the spring design, and tested
• For the customer, a completed and tested analysis tool which we will help integrate
with their own 6-DOF model if needed
• Panel 2, the windward panel, poses greatest risk of recontact and will be the focus of
this presentation
1/5/2015
Introduction
This Analysis Draws Upon Prior Experience for this Class of Problem:
• Hall, L.H., Eppard, W.M., Applebaum, M.P., Mitchell, C. R., “Computational Simulation
Techniques of Panel Fairing Jettison from a Launch Vehicle System”. JANNAF 959,
2008.
o Comparisons between fully time accurate and quasi-unsteady
• Hall, L.H., Applebaum, M.P., Eppard, W.M.,"Debris Transport Modeling Techniques on
Launch Vehicle Systems",49th AIAA Aerospace Sciences Meeting, Orlando, Fl, Jan 4-
7, 2011.
o Comparisons between fully time accurate and quasi-unsteady
• Hall, L.H., Parthasarathy, V., “Validation of an Automated Chimera/6-DOF
Methodology for Multiple Moving Body Problems”, AIAA 98-0753, 1998
1/5/2015
Bounding the Database
• Provided bounding trajectory points from GNC:
Mach
number 5.99 5.53 5.85 5.9 8.55 7.45 8.48 8.32
Angle-of-
attack, deg −20.6 −20.75 −20.85 −20.43 −15.92 −15.01 −15.81 −15.09
Sideslip
angle, deg −4.51 −4.78 −4.77 −5.01 3.6 3.72 3.45 3.21
• From these points we determined the bounding flight conditions.
• It was noted a clear bias of beta associated with the alpha and Mach
number.
• Was it worth the computational space to model positive beta at the
lower Mach or negative beta at higher Mach? Could those
conditions ever exist?
1/5/2015
Bounding the Database
• Two choices were considered:
1) Construct the database in a traditional manner, bounded by
alpha/beta. Due to time constraints, a single Mach number
would be chosen. It was determined the single Mach number
would split the bounds provided by GNC; Mach = 7.0.
2) Produce a database with associated alpha/beta tied to Mach,
and simulate results at the bounding end of the Mach
numbers, approximately 5.5 and 8.5. For the same number of
stations in time, this requires exactly half the simulations.
• Which would be the more dominant effect? Flow turning due to
the Mach number variation, or small variations of alpha/beta?
o From 2D shock tables, we might expect a differential of
shock wave angle on the order of several degrees for the
bounding Mach numbers.
1/5/2015
Bounding the Database
•It was decided to have a quick look at Mach number versus Alpha/Beta effects:
Cases 1,2: Panels open at 30 degrees, Mach 5.4, alpha -21, beta -5 and +4
Cases 3,4: Panels open at 60 degrees, Mach 5.4, alpha -21, beta -5 and +4
Cases 5,6: Panels open at 30 degrees, Mach 8.55, alpha -21, beta -5 and +4
Cases 7,8: Panels open at 60 degrees, Mach 8.55, alpha -21, beta -5 and +4
•The results show that for the most windward panel, the most significant Mach number
effect between Mach 5.4 and Mach 8.55 produced a moment difference of 5.9% seen at
30 degree panel rotation. The maximum beta effect on panel moment was
coincidentally also 5.9%, but was observed at 60 degree panel rotation.
•It should be noted the beta percentage difference is obtained over the full range of
beta, a range that should not exist at a given Mach number. So ultimately, it appears
the Mach number effect could be more significant for expected range of possible
conditions. It appears the untraditional approach of Mach tied to alpha/beta is a viable
and cost-efficient choice that would provide roughly the same accuracy as modeling
alpha/beta variation with half the simulations.
1/5/2015
Bounding the Database
• HOWEVER: the limitation of a chosen Mach number being tied to alpha/beta seems
like a restrictive element of the database that could cause the database to lose
relevance it might otherwise maintain if dispersed conditions change. It was
determined that time existed for the larger computational matrix to be obtained by
delivery date, so that option was chosen.
• The chosen bounds for the database were:
1/5/2015
Panel Geometry
1/5/2015
Panel Orientation
1/5/2015
Database Zones
Database will Consist of Three Zones:
1) Panel on the Hinge: considered most crucial in trajectory determination. Panel-
to-panel effects captured
2) Near the body: where body proximity effects are taken into account
3) Far from the body: panel alone data utilized in this region where body proximity
effects considered small or unimportant
1/5/2015