Knowledge base

Structural analysis,
in plain terms.

25 short, accurate answers to the aerospace stress-analysis questions I get asked most — search them or jump by topic.

25 answers

Damage Tolerance & Fatigue

What is the difference between slow crack growth and fail-safe?
Both are damage-tolerance design philosophies recognised under FAR/CS 25.571. Slow crack growth assumes a crack exists in a single, non-redundant load path; you demonstrate it grows slowly and predictably and is found at scheduled inspection before reaching critical length. Fail-safe instead relies on redundancy or crack-arrest features so that if one element fails, load redistributes and the remaining structure still carries a defined residual-strength load to the next inspection. The choice drives different critical locations, load cases and inspection programmes, so it must be made before the analysis — not after. Read the full article →
What is damage tolerance in aircraft structures?
Damage tolerance is the discipline of assuming a structure already contains a flaw — a manufacturing defect or a fatigue crack — and demonstrating the aircraft remains safe anyway. Safety is shown either through predictable crack growth plus scheduled inspection, or through redundancy and residual strength. It replaced the older safe-life-only mindset after service failures showed that real structures crack, and the question is not whether but how you manage it. Read the full article →
What does DaDT mean?
DaDT stands for Durability and Damage Tolerance. Durability (fatigue) covers how long a structure lasts before cracks initiate under repeated load; damage tolerance covers how the structure behaves once a crack exists. A DaDT engineer owns both halves: predicting crack initiation lives and substantiating slow crack growth or fail-safe behaviour, including setting inspection intervals.
What is the Paris law in crack growth?
The Paris law is the basic relation governing fatigue crack growth: da/dN = C·(ΔK)^m, where da/dN is the crack growth per cycle, ΔK is the stress-intensity-factor range, and C and m are material constants. Because the exponent m is typically 2–4, growth scales steeply with ΔK — a large cycle does far more damage than its count suggests. It describes the stable mid-range of growth; it does not capture threshold behaviour at low ΔK, fast fracture near critical K, or load-interaction (retardation) effects.
What is spectrum truncation and omission in fatigue analysis?
A fatigue load spectrum is edited at both ends before analysis. Omission deletes very small low-amplitude cycles that contribute negligible damage, to make the run tractable. Truncation clips rare high loads to a ceiling. Both change the predicted life: omit too aggressively and you lose cycles that matter in aggregate near a notch; truncate the real overloads out and you can lose the crack-growth retardation they would have caused. The levels must be justified and documented, because the conservative direction is not obvious. Read the full article →
What is retardation (load interaction) in crack growth?
Retardation is the slowing of fatigue crack growth that follows a tensile overload. The overload creates a large plastic zone at the crack tip; as the crack grows into the compressive residual stresses left behind, growth slows — sometimes for thousands of cycles, occasionally arresting entirely. The constant-amplitude Paris law ignores this, so a retardation model (Wheeler, Willenborg, or a crack-closure approach after Elber) is needed for variable-amplitude spectra. Crediting retardation you cannot substantiate by test is borrowing against a life you have not earned. Read the full article →
What is the difference between safe-life and damage-tolerant design?
Safe-life design retires a part after a demonstrated number of cycles, assuming it is crack-free throughout that life with a scatter factor applied — common for landing gear and rotating parts. Damage-tolerant design instead assumes a flaw is present from day one and manages it by inspection or redundancy. Safe-life is simpler but wastes life and offers no protection against unexpected defects; damage tolerance is the default for fixed-wing primary structure. Read the full article →

Finite Element Analysis

What is GFEM in aircraft stress analysis?
GFEM stands for Global Finite Element Model — a coarse, airframe-wide model whose job is to bookkeep the load path, distributing internal loads such as axial, shear and bending across spars, webs, skins and stringers. Its local stresses at fillets, cut-outs or holes are not trustworthy because the mesh is deliberately coarse. Those detailed stresses come from sub-models driven by GFEM grid-point forces. In short, you trust a GFEM for loads, not for local stress. Read the full article →
Why is mesh convergence important in FEA?
A finite element result is only meaningful if it stops changing as the mesh is refined — that is mesh convergence. It confirms the reported quantity reflects the physics rather than discretisation error, which matters most at stress concentrations where coarse meshes under-predict peaks. The right practice is to converge the quantity that goes into the margin (often a section force or a stress at a defined offset), check averaged-versus-unaveraged stress recovery, and compare against a hand calculation. Convergence of *a* number is not convergence of *the* number you should be using. Read the full article →
Why do stresses at sharp corners never converge (singularities)?
At a re-entrant corner with zero fillet radius, a point load, or a single-node constraint, the idealised mathematical stress is infinite. The FE solution cannot show infinity, so the peak stress simply climbs every time you halve the element — slowly and forever. Refining the mesh there is the wrong fix; the answer is to model the real fillet radius, smear the load over its true footprint, or read the convergent section force instead of the divergent peak stress. Read the full article →
What is sub-modelling and how do you go from global to detail?
Sub-modelling carries loads from a coarse global model into a fine, geometrically detailed local model that actually resolves the stress. You cut the boundary far enough from the feature of interest (Saint-Venant: at least one characteristic dimension away), apply the GFEM grid-point forces, and — crucially — reconcile the loads crossing the cut against the global model before trusting anything. A detailed model that quietly disagrees with the global load path is not more accurate; it is a different problem wearing the same geometry. Read the full article →
When do you need nonlinear FEA instead of linear?
Linear analysis assumes small displacements, linear-elastic material and unchanging boundary conditions — fine for most strength sizing. You need nonlinear analysis when any of those break: large deflections (geometric nonlinearity), material yielding or plasticity, or contact and changing load paths such as bolted joints, snap-through and post-buckling. Nonlinear runs are more expensive and need careful convergence control, so you reserve them for where the physics genuinely demands them.

Materials & Allowables

What is the difference between A-basis and B-basis allowables?
Both are statistically derived material strength values at the same 95% confidence. A-basis is the value exceeded by at least 99% of the population; B-basis by at least 90%. A-basis is required for single, non-redundant load paths where one failure is catastrophic; B-basis is allowed where load can redistribute (fail-safe structure). The basis is therefore a structural decision driven by your damage-tolerance philosophy, not a materials one — choosing A where B would do simply costs mass. Read the full article →
What is a knockdown factor?
A knockdown factor reduces a pristine, room-temperature material allowable to account for the real environment a part lives in: elevated temperature and absorbed moisture (hot/wet), open holes, impact damage, fatigue scatter and so on. For composites especially, these stack — open-hole compression after hot/wet conditioning can sit at a small fraction of the unnotched laminate strength. The honest working allowable is the pristine value with every applicable knockdown applied, and each one should be visible in the margin. Read the full article →
What is MMPDS and where do metallic allowables come from?
MMPDS (Metallic Materials Properties Development and Standardization) is the FAA-recognised handbook of statistically substantiated mechanical properties for aerospace metals — the successor to MIL-HDBK-5. It provides A-basis and B-basis strengths, moduli and fatigue data derived from large qualification test datasets. Stress engineers pull allowables from it rather than re-deriving them, but you still have to choose the right basis, temperature and condition for your detail. Read the full article →
Why is hot/wet the critical condition for composites?
Carbon-fibre composites absorb moisture, and at elevated temperature that moisture plasticises the polymer matrix. Because matrix-dominated properties — compression, shear and bearing — degrade more than fibre-dominated tension, the hot/wet corner of the environmental envelope usually governs compression-driven sizing. That is why composite allowables are qualified at the environmental extremes and the worst case is carried, while cold/dry can instead govern matrix tension or metallic toughness. Read the full article →

Joints & Details

What is bearing-bypass in a bolted joint?
In a loaded bolted joint, the load at each hole splits two ways: part transfers through the fastener into the plate as bearing, and the remainder passes around the hole on its way to the next fastener as bypass. The bearing-bypass ratio — not just the total load — governs the local stress, net-section check and fatigue life at the hole. Because end fasteners in a row carry more than their fair share, the distribution is checked with a flexibility method (Huth, Swift) and confirmed against FE rather than assumed uniform. Read the full article →
What is a stress concentration factor (Kt)?
Kt is the ratio of the peak local stress at a geometric feature — a hole, fillet or notch — to the nominal far-field stress. For a small circular hole in an infinite plate under tension the classic elastic value is Kt ≈ 3 on the gross stress (the Kirsch solution). It is meaningless until you state the reference area (gross or net), and as a hole grows relative to width the net-section Kt actually drops while the net stress rises. It is a theoretical elastic factor — the starting point for both static sizing and fatigue assessment. Read the full article →
What is the fatigue notch factor (Kf)?
Kf is the effective stress concentration for fatigue, always smaller than the elastic Kt: Kf = 1 + q·(Kt − 1), where q is the notch-sensitivity factor between 0 and 1. Physically, fatigue damage responds to stress averaged over a small process-zone of material, not the mathematical point at the notch root, so a very sharp notch does less fatigue damage than its elastic Kt implies. Sharp notches in high-strength alloys approach q = 1 (Kf ≈ Kt); blunt notches in ductile, lower-strength materials sit lower. Read the full article →

Certification & Standards

What is EASA CS-23?
CS-23 is the European Union Aviation Safety Agency certification specification for normal-category aeroplanes — light aircraft and trainers — equivalent in scope to FAA FAR Part 23. It defines the structural strength, fatigue and damage-tolerance requirements an airframe must be substantiated against, along with the load cases and factors of safety. A structural substantiation programme produces stress reports that demonstrate compliance with the relevant CS-23 paragraphs to the certification authority.
What is the difference between limit load and ultimate load?
Limit load is the maximum load expected in service; the structure must carry it with no detrimental permanent deformation. Ultimate load is limit load multiplied by the factor of safety — typically 1.5 for aircraft — and the structure must withstand it without failure, though permanent deformation is allowed. Static substantiation checks both: no yielding at limit, no rupture or collapse at ultimate.
What is a margin of safety (reserve factor)?
The margin of safety, MS = (allowable / applied) − 1, expresses how much strength is left over once the applied stress is compared to the allowable at the relevant load level; a positive margin means the part passes, zero means it is exactly sized. The closely related reserve factor is simply allowable / applied. The number is only as honest as the allowable behind it — which basis, which environment, which knockdowns — so a margin should always carry that provenance.

Methods, Tools & Career

What software do aerospace stress engineers use?
The core solvers are MSC Nastran, Abaqus and ANSYS Mechanical, with Altair OptiStruct for optimisation; pre/post-processing is done in MSC Patran, Femap or Altair HyperMesh, and geometry in CATIA V5. Beyond the GUI, much daily work is automation: Python (with libraries like pyNastran), and scripting in Nastran DMAP, Patran PCL or HyperMesh Tcl. Hand-calculation workbooks in Excel/VBA remain essential for joints, lugs and quick checks. Read the full article →
What is pyNastran and why automate stress analysis?
pyNastran is an open-source Python library for reading and writing MSC Nastran input decks and result files (.op2 / .f06), letting you pull grid-point forces, element loads and stresses programmatically. Automating the pipeline — from results to margin-of-safety tables and report figures — removes hours of manual bookkeeping and, more importantly, makes the engineering assumptions visible: the faster the arithmetic runs, the harder it is to hide which allowable, load case or knockdown was used. Read the full article →
How do you become an aerospace stress / DaDT engineer?
The usual path is a mechanical or aerospace engineering degree with a structures focus, often followed by an M.Sc. in solid mechanics, then learning the trade on real programmes under experienced engineers. Fluency in finite element analysis, classical hand calculations and the canon (Bruhn, Niu, Megson, Schijve) matters as much as any single tool, and certification programmes teach you to write substantiation reports that an authority will accept. Fatigue and damage tolerance is a specialism you grow into — it rewards judgement built over years, because the assumptions, not the arithmetic, are where the difficulty lives.

Have a structural-analysis question that isn't here?

Ask me directly