FEM software — NASTRAN, ABAQUS (concepts and use)

3.6.20 · Physics › Spacecraft Structures & Systems Engineering

FEM Software Kya Karta Hai

Core Workflow

Pre-processing: Geometry banao, materials define karo, mesh apply karo (structure ko elements mein divide karo)
Solver: Stiffness matrices assemble karo, boundary conditions apply karo, $Ku = F$ solve karo
Post-processing: Stress contours, deformation, safety margins visualize karo

Ye structure kyun? Kyunki physics (elasticity PDEs) → discretization (FEM) → linear algebra (solve) → engineering decisions (interpret) alag-alag concerns hain.

NASTRAN: Aerospace ka Workhorse

NASTRAN (NASA Structural Analysis) 1960s mein Apollo program ke liye develop kiya gaya tha. Philosophy: fast, reliable, linear analysis bade aerospace structures ke liye. Text-based input deck (.bdf Bulk Data File), proven algorithms, extensive validation heritage.

Spacecraft engineers ise kyun pasand karte hain:

Launch vehicle qualification: dynamic loads ke liye proven (random vibration, shock, acoustic)
Modal analysis: natural frequencies dhundta hai resonance avoid karne ke liye
Linear = fast: 1 million DOF models ko minutes mein analyze karo

Key Capabilities

Analysis Type	Kya Karta Hai	Kab Use Karte Hain
Linear Static	$Ku = F$ , displacements/stresses dhundta hai	Steady loads (gravity, thermal)
Modal	Eigenvalue problem $(K - ω^{2} M) ϕ = 0$	Vibration modes dhundna, resonance avoid karna
Frequency Response	Har frequency $ω$ pe solve karo: $(- ω^{2} M + iω C + K) u = F (ω)$	Harmonic forcing (rotating machinery)
Random Vibration (PSD)	Random loads ka statistical response	Launch vibration environments

Problem: Ek L-bracket (aluminum, 100mm × 50mm × 3mm) ek end pe bolted hai. Pehle 3 natural frequencies dhundo.

NASTRAN Input Deck Structure:

$ Executive Control
SOL 103          $ Modal analysis
$ Case Control
METHOD = 1       $ Eigenvalue method
SPC = 10         $ Boundary conditions
$ Bulk Data
EIGRL,1,,,10     $ Lanczos, find 10 modes
GRID,1,0,0,0   $ Node 1 at origin
...
CQUAD4,1,1,2,3,4  $ Quad element
MAT1,1,70.E9,,2700  $ E=70GPa, rho=2700kg/m³
SPC1,10,123456,1    $ Fix all DOF at node 1

Ye cards kyun?

SOL 103: Modal analysis ke liye solution sequence
EIGRL: Eigenvalue solver (Lanczos method bade sparse matrices ke liye)
CQUAD4: 4-node shell element (thin structures ke liye acha)
MAT1: Isotropic material (Young's modulus, density)
SPC1: Single-point constraint (boundary condition)

Interpretation: Output dikhata hai $f_{1} \approx 850$ Hz (first bending mode). Agar launch vibration mein 850 Hz pe energy hai, toh stiffener ya damper add karo.

ABAQUS: Nonlinear Specialist

Jab Linear Kaafi Nahi Hota

ABAQUS (Dassault Systèmes se) nonlinearities handle karta hai:

Material nonlinearity: Plasticity, hyperelasticity (rubber seals), composite damage
Geometric nonlinearity: Large deformations (deployable booms), buckling
Contact: Bolted joints, friction, separation

Spacecraft ko ye kyun chahiye:

Composite layup analysis: delamination, fiber failure modes
Deployable mechanisms: hinges, cables, large rotations
Crush simulations: landing gear impact

Example: Composite Panel Progressive Failure

Problem: Ek carbon fiber panel (8 plies, quasi-isotropic layup) compression ke under. Ye kab fail karta hai?

ABAQUS Approach:

Material model: Hashin damage criteria (fiber tension/compression, matrix cracking)
Elements: Continuum shell elements layup definition ke saath
Analysis: Static Riks (arc-length method post-buckling ke liye)
Output: Load-displacement curve, damage initiation locations

Ye step-by-step kyun?

Step 1: Hashin fiber/matrix mein stress alag-alag check karta hai: $(\frac{σ _{11}}{X _{T}})^{2} + α (\frac{τ _{12}}{S})^{2} \geq 1$ (fiber tension failure)
Step 2: Har ply ek alag layer hai apne orientation ke saath (0°, 45°, -45°, 90°)
Step 3: Riks method equilibrium path follow karta hai peak load ke baad bhi (snap-through)
Step 4: Color contour "failure index" = 1.0 dikhata hai jahan damage shuru hota hai

Result: Panel 12.3 kN pe fail karta hai (ply interfaces pe matrix cracking), phir 14.1 kN pe catastrophic fiber breakage. Design margin insufficient hai—plies add karo ya layup change karo.

Worked example Bolted Joint Contact Analysis

Setup: Do aluminum plates (5mm thick) M6 bolt se joined, 10 N·m pe torqued. Contact pressure distribution aur slip dhundo. ABAQUS Input:

*SURFACE INTERACTION, NAME=CONTACT: Friction coefficient μ = 0.2 define karo
*CONTACT PAIR: Master/slave surfaces specify karo (bolt head/plate)
*STEP, NLGEOM=YES: Nonlinear geometry (bolt preload stiffness change karta hai)
*CLOAD, OP=NEW: Bolt preload equivalent force ke roop mein apply karo $F = \frac{T}{0.2 d} \approx 8333$ N Har setting kyun?
Friction matters hai: iske bina plates freely slide karti hain (unphysical)
NLGEOM=YES: bolt tension joint ko stiffen karta hai (geometric nonlinearity)
Preload formula: $T = k F d$ , jahan $k \approx 0.2$ lubricated threads ke liye, $d = 6$ mm Output: Contact pressure 180 MPa pe peak hoti hai (bearing strength ke andar), 5 kN shear load ke under koi slip nahi. Joint safe hai.

Practical Workflow Differences

| Aspect | NASTRAN | ABAQUS | |-----|------| | Input | Text deck (.bdf) | Text deck (.inp) ya CAE GUI | | Speed | Linear ke liye fast (optimized solvers) | Slow (iterative nonlinear) | | Typical Model Size | 1-10M DOF | 10k-1M DOF (nonlinear overhead) | | Best For | Spacecraft primary structure, launch loads | Composites, deployables, detailed joints | | Validation | 50+ saal ka flight heritage | Crash/impact ke liye industry standard |

Text decks kyun? Version control, automation (parametric studies), reproducibility. GUI sikhne ke liye acha hai, lekin production analysis scripts use karta hai.

Common mistake Common Pitfall: FEM Results pe Blindly Trust Karna

Galat approach: Analysis run karo, stress = 200MPa < 250 MPa yield dekho, safe declare karo. Ye sahi kyun lagta hai: Software ne ek number diya, aur wo limit se neeche hai.

Problem:

Mesh bahut coarse ho sakta hai (stress concentrations miss ho jaati hain)
Boundary conditions galat ho sakti hain (fixed vs. flexible support)
Load case incomplete ho sakta hai (thermal expansion + vibration bhool gaye)
Element type inappropriate ho sakta hai (thin shell ko bricks se model kiya → shear locking) Fix: Convergence study (mesh refine karo jab tak results 5% se kam change hon), sanity checks (simple geometry ke liye hand calc), validation (test data se compare karo), sensitivity analysis (uncertain inputs vary karo). FEM ek tool hai, oracle nahi.

Software Selection Decision Tree

NASTRAN kab use karein:

Linear analysis (adhiktar spacecraft structures linear elastic hote hain)
Bade models (millions of DOF)
Frequency domain (random vibration, acoustic)
Heritage/certification requirements (NASA, ESA standards)

ABAQUS kab use karein:

Material nonlinearity (composites with damage, plasticity)
Geometric nonlinearity (post-buckling, large rotations)
Contact problems (bolted joints, seals, deployables)
Explicit dynamics (impact, shock, crash)

Dono kab use karein:

Global model NASTRAN mein (linear, fast) → forces extract karo → ABAQUS mein local detailed model (nonlinear joint)

Recall 12 Saal ke Bacche ko Samjhao

Socho tum ek treehouse bana rahe ho. Tum jaanna chahte ho: Kya floor mera weight hold karega? Kya hawa ise giraa degi?

Tum ise bana ke dekh sakte ho kya toota—lekin isse wood barbad hoti hai aur ye dangerous bhi hai. Iske bajaye, tum ek computer model banate ho: treehouse draw karo, computer ko batao ye wood ka bana hai (jo force ke under thoda jhukta hai), aur kaho "mera weight 50 kg hai aur hawa 60 km/h pe push karti hai."

Computer treehouse ko tiny pieces mein chop karta hai (jaise LEGO bricks), figure out karta hai har piece kaise jhukta hai aur apne neighbors pe kaise push karta hai, aur total calculate karta hai. Ye tumhe batata hai: "Floor 2 cm jhukta hai (safe!) lekin left wali rope breaking strength ka 90% pe hai (ek aur rope add karo!)."

NASTRAN ek super-fast calculator ki tarah hai jo simple materials (wood, metal) aur chote jhukao ke liye perfect hai. ABAQUS ek scientist ki tarah hai jo weird stuff handle kar sakta hai: rubber bands jo bahut stretch hote hain, pieces jo crash hote hain, ya materials jo crack karte hain. Spacecraft engineers dono use karte hain: NASTRAN badi structure ke liye, ABAQUS tricky parts ke liye (hinges, seals, composites).

Verification and Validation

Standard checks:

Equilibrium: $\sum F = 0$ , $\sum M = 0$ (reactions applied loads balance karte hain)
Energy: Strain energy $U = \frac{1}{2} u^{T} Ku$ work $W = F^{T} u$ ke barabar honi chahiye
Symmetry: Agar geometry/load symmetric hai, toh deformation bhi honi chahiye
Limiting cases: Slender beam → Euler-Bernoulli formula, thin plate → Kirchhoff theory

Spacecraft Design se Connections

Structural Analysis Methods: FEM ek method hai; analytical solutions, testing se compare karo
Vibration and Modal Analysis: NASTRAN modal = $ω_{n}$ dhundo Launch Vehicle Loads ke saath resonance avoid karne ke liye
Composite Materials: Carbon Fiber Structures ke liye ABAQUS layup analysis
Thermal-Structural Coupling: NASTRAN temps export karo → thermal stresses (ya coupled analysis)
Stress Analysis and Margins: FEM stresses output karta hai → Factors of Safety apply karo → design margins
Model Correlation: FEM predictions ko Ground Test Procedures data se match karo

#flashcards/physics

Linear FEM mein solve kiya jaane wala fundamental equation kya hai? :: $Ku = F$ , jahan $K$ global stiffness matrix hai (element stiffness matrices se assembled $k_{e} = \int B^{T} DB d V$ ), $u$ displacement vector hai, aur $F$ applied load vector hai. Ye discretized equilibrium represent karta hai.

NASTRAN spacecraft primary structures ke liye kyun prefer kiya jata hai?

Fast linear analysis validated solvers ke saath, bade DOF models ke liye excellent (1-10M), launch loads aur frequency-domain analysis ke liye proven heritage (modal, random vibration), certification ke liye industry-standard (NASA, ESA).

ABAQUS ka Newton-Raphson solver linear FEM se alag kya karta hai?

Iteratively

R (u) = F_{ext} - F_{int} (u) = 0

solve karta hai tangent stiffness

K_{T}

compute karke,

K_{T} Δ u = R

solve karke, aur convergence tak

u

update karke. Nonlinear material/geometry/contact handle karta hai jahan stiffness displacement pe depend karti hai.

Mesh convergence study kya hai aur ye kyun critical hai?

Analysis progressively finer mesh ke saath rerun karo (2×, 4× elements) jab tak key outputs (peak stress, max displacement) 5% se kam change hon. Critical isliye kyunki FEM ek approximation hai—bahut coarse mesh stress concentrations miss karta hai, deformations underestimate karta hai. Ye verification step hai, validation nahi.

Teen nonlinearities name karo jo NASTRAN ki jagah ABAQUS require karti hain :: (1) Material nonlinearity: plasticity, composite damage, hyperelasticity; (2) Geometric nonlinearity: large deformations, post-buckling, snap-through; (3) Contact: friction, separation, bolted joints. NASTRAN sirf linear elastic hai.

Element stiffness matrix first principles se kaise derive karte hain?

Virtual work se shuru karo:

\int_{V} σ : δ ϵ d V = δ W

. FEM discretization substitute karo:

u = N u_{e}

(shape functions),

ϵ = B u_{e}

(strain-displacement),

σ = D ϵ

(Hooke's). Result:

k_{e} = \int_{V} B^{T} DB d V

Modal analysis eigenvalue problem kya hai?

(K - ω^{2} M) ϕ = 0

, jahan

K

stiffness hai,

M

mass matrix hai,

ω

natural frequencies hain (rad/s), aur

ϕ

mode shapes hain. Launch vibration environments ke saath resonance avoid karne ke liye frequencies dhundhta hai.

GUI available hone ke bawajood text input decks kyun standard rehte hain?

Version control (git mein changes track karo), automation (scripts ke zariye parametric sweeps), reproducibility (exact rerun), collaboration (code review), optimization tools ke saath integration. GUIs sikhne ke liye aache hain, decks production ke liye.