Analyze/Size Spacecraft Structures: Loads Development & Structural Dynamics

[Dikuza]

Hello, I'm a mechanical/aerospace design engineer trying to expand into structural analysis.

What I'm looking for:
* A guide or summary as to 1) how spacecraft design loads are determined, and 2) how components are analyzed to these loads
* Recommended course to learn the above


Ideally, I'd like to see some sort of table that summarizes all the different loading conditions.
I've started the table below as an example where the first row is information taken mostly from Space Mission Analysis and Design by Wiley Larson pg 461: "Random vibration from engines and other sources is a critical source of load. At lift-off, the major source of random vibration is acoustic noise (pressure waves reflecting off the launch pad and service tower), which radiates from the engines to engulf the vehicle. Acoustics develop from aerodynamic turbulence when the vehicle passes through the transonic portion of its flight. Structures with high surface area and low mass, including skin sections and solar array panels, respond strongly to acoustic noise."

Other than this row, however, is information that I've filled in based off my limited knowledge (please correct any mistakes) and have no idea where things like steady-state acceleration, pogo vibration, attitude maneuvers, or sinusoidal vibration would fit in or what they affect.

I understand it's probably not reasonable to boil down such a complex subject into a single table but I think would be a great summary in starting to understand the different load cases, where they come from, what they affect, and how they're analyzed.
From there I'd like to learn how these design loads (to be used in analyses) are actually approximated. For a satellite, I know this mostly comes from the launch vehicle provider's coupled-loads analysis (CLA), but what about for a human vehicle? 
The bottom two highlighted rows are the load cases that I assume would drive the design of primary structure of a human-rated return spacecraft (Dragon, Starliner, Orion) of which I'm most interested.



Where I've looked:
Besides the above mentioned text from Larson, I'm also aware of and in the process of reading:
* NASA-STD-5002 Load Analyses of Spacecraft and Payloads - this describes the overall process of load cycle analysis (LCA) but not very detailed or how to actually approximate loads.
* Structural Design and Analysis for Aerospace Engineers Course Book by Thomas P. Sarafin - this has a lot of great info but focuses on launch vehicles and doesn't mention human spacecraft much.
* Spacecraft Structures and Mechanisms: From Concept to Launch by Thomas P. Sarafin - I came across this referenced elsewhere and have not been able to find an online pdf version yet. Will consider buying hard copy.
* Manned Spacecraft Design Principles by Pasquale Sforza (2016) - just started reading this and seems like a great reference in general but the section for structural dynamics is very short.

Any help is much appreciated! Thanks for reading.

[laszlo]

A human launch vehicle is a satellite with meat in it, so the LV provider's CLA is completely valid there, too. You just have to worry about different limits.

Same for the shock load for any vehicle meant to return in one piece. You just take that analysis and use the limits required by people instead of mechanical structures.

For the splashdown/landing entry I'd have something about sizing wings or parachutes. Parachutes also typically fire mortars to deploy them so there's that shock to consider, as well as the chutes inflating. The interesting thing about chutes is that they're flexible structures mostly doing everything in tension and torsion without much compressive load like primary structure.

Do keep us updated on your progress. This sounds potentially like a textbook in the making.

[jarmumd]

LV Loads and Dynamics can be a deep and complicated subject, but it can be roughly explained at a high level.  Note that Loads and Dynamics can mean different things for different programs, but I'll explain my experience with large vehicles like SLS and Stratolaunch.

There are two fundamental steps, which have direct analogs to classic beam bending theory in Finite Element Analysis.  So if you were to evaluate a 10 element cantilever beam, Step 1 would be to apply forces either statically or dynamically (depending on what you need as output), keeping in mind that reactions are just forces that exactly balance out the internal (inertial) and external (applied) forces.  This step calculates the resultant displacement of the FEM, due to the load conditions.  Step 1 computes displacements, so to get the "loads", you apply the inverse of the stiffness matrix at specific section cuts, resulting in interface forces between elements, and finally in a shear and bending moment diagram.  Step 2 is to take those elemental shear and bending moments (really in this simple case, it's still displacements) and determine the stresses between them using the element stress equations.

So for a LV, we have methods to create an "equivalent beam" model from a high fidelity FEM, apply Loads in the form of Pseudo-static thrust loads, Static Aeroelastic Loads, Transient Tran-sonic Buffet, Transient Liftoff Loads, etc...  Often these are separate analysis.  Then compare all of those shear and bending moment results to come up with an overall max load.  Then verify that the LV design has higher margins than the applied loads.  That last step typically involves a stress analyst taking a table of results from a Loads Data Book, and then applying them to a very high fidelity stress model of some sub-section of the LV.  This is why we often have "loads" fems, which are low fidelity but are easily tuned to have the correct modes and mode shapes, and "stress" fems which are incredibly detailed, but difficult to tune and take too long to run for dynamics.

Very generally speaking, Max Q-Alpha (maximum dynamic pressure at the maximum Angle of Attack) is the dominant load and drives vehicle design.  But, each regime can drive a peak load for a sub-section or a component.  For instance, Max Q-Alpha will not drive bending moments at the base of the LV, but preload wind loads can.

This is just the tip of the iceberg.  But you don't see much about it, because it is very specialized (typically taught in house), and there are often 10 stress and design engineers for 1 loads analyst.

[Jim]

Staging and spacecraft deployment are shock loads due to release mechanisms

Launch abort would drive loads (high acceleration)
Landing would also be loads.  Shock is a shorter high intensity event.  Like hitting a bell with a hammer.
Engine ignition or shutdown is also a load event.

Random vibe is from engine vibration or acoustic noise.

Random vibe testing is more for workmanship validation

Modal surveys are used to validate spacecraft structural models use in CLAs