Composite materials — fiber-matrix, ply properties, laminate theory: Question bank

Before the traps, a short glossary so every symbol below is earned before use.

Definition Every symbol used on this page

==Fiber volume fraction $V_{f}$ ==: the fraction of the composite's volume that is fiber (e.g. $V_{f} = 0.6$ means 60% fiber). ==Matrix volume fraction $V_{m}$ == is the rest, so $V_{f} + V_{m} = 1$ .
== $E_{f}$ , $E_{m}$ ==: the stiffness (Young's modulus) of the pure fiber and pure matrix on their own — carbon fiber $E_{f} \sim 230$ GPa, epoxy $E_{m} \sim 3$ GPa.
== $E_{1}$ , $E_{2}$ ==: the composite's stiffness along the fibers (axis 1) and across them (axis 2).
==Strain $ϵ$ ==: how much a piece stretches, as a fraction of its length; $ϵ_{f}$ and $ϵ_{m}$ are the stretches of the fiber part and matrix part.
==Shear modulus $G_{12}$ ==: stiffness against sliding (angling) rather than stretching, in the 1–2 plane.
==Poisson ratio $ν_{12}$ (major)==: pull along axis 1 and the material squeezes in axis 2 — $ν_{12}$ is how much. == $ν_{21}$ (minor)== is the same coupling referenced to pulling along axis 2, and $ν_{21} = ν_{12} E_{2} / E_{1}$ .

Recall The two directions every answer refers to

What do the material axes "1" and "2" mean? ::: Axis 1 runs along the fibers (stiff, strong); axis 2 runs across the fibers, through the weak matrix — the whole point of a ply is that these two directions behave completely differently.

Two pictures anchor the two rules and the four matrices you'll meet below.

Figure — Composite materials — fiber-matrix, ply properties, laminate theory

True or false — justify

TF1. A composite with

E_{1} = 140

GPa is "as good as" a 140 GPa isotropic metal.

False — that 140 GPa only exists along the fibers; across them

E_{2}

may be ~10 GPa, so the material is roughly 14× softer transversely and cannot be treated like isotropic aluminium.

TF2. The longitudinal Rule of Mixtures

E_{1} = E_{f} V_{f} + E_{m} V_{m}

assumes fiber and matrix feel the same stress.

False — it assumes the same strain (they are bonded and stretch together in parallel, like parallel springs); each phase then carries stress proportional to its own stiffness.

TF3. The transverse modulus

E_{2}

is close to the average of

E_{f}

and

E_{m}

False —

E_{2}

uses the inverse (series) rule and is dominated by the weakest link (the matrix), so

E_{2}

sits near

E_{m}

, not near the average.

TF4. Because

ν_{21} = ν_{12} E_{2} / E_{1}

, the minor Poisson ratio

ν_{21}

is smaller than the major

ν_{12}

True — since

E_{2} < E_{1}

, the factor

E_{2} / E_{1} < 1

, so

ν_{21} < ν_{12}

; both describe the same physical coupling, just referenced to different loading directions.

TF5.

Q_{11}

is exactly equal to

E_{1}

False —

Q_{11} = E_{1} / (1 - ν_{12} ν_{21})

is slightly larger than

E_{1}

because Poisson coupling stiffens the constrained response, though the difference is usually under 1%.

TF6. Adding more fiber (

V_{f} \to 1

) always makes a better composite.

False — beyond ~0.65 there isn't enough matrix to wet and support the fibers, so voids form, load transfer fails, and strength actually drops.

TF7. A single unidirectional ply loaded off-axis develops shear even under pure tension.

True — once fibers sit at an angle

θ

, a tensile pull resolves into components along and across the fibers, and the transformed stiffness

[\overset{ˉ}{Q}]

gains extension–shear coupling terms.

TF8. Fatigue resistance is a selling point because composites, unlike metals, tend not to propagate cracks the same way.

True — there is no single crystal-lattice crack marching through; damage spreads as distributed fiber/matrix micro-events, giving much flatter fatigue curves (relevant to launch load cycling).

Spot the error

SE1. "In the series (transverse) model I set

ϵ_{f} = ϵ_{m}

because both phases stretch the same."

Wrong — series loading means equal stress, not equal strain; the softer matrix strains more (

ϵ_{m} > ϵ_{f}

), and the total strain is the volume-weighted sum of the two.

SE2. "

1/ E_{2} = V_{f} / E_{f} + V_{m} / E_{m}

, so with

V_{f} = 0.6

the fiber dominates

E_{2}

Wrong — in an inverse sum the largest denominator contributes the smallest term, so the tiny

V_{m} / E_{m}

term (small

E_{m}

) dominates

1/ E_{2}

; the matrix, not the fiber, controls transverse stiffness.

SE3. "The stiffness matrix

[Q]

has a nonzero

Q_{16}

term coupling normal stress to shear."

Wrong — in the material axes the ply is orthotropic, so

Q_{16} = Q_{26} = 0

; those coupling terms only appear in the transformed

[\overset{ˉ}{Q}]

after rotating to an angled layup.

SE4. "To go from compliance

[S]

to stiffness I just take reciprocals of each entry."

Wrong —

[Q] = [S]^{- 1}

is a full matrix inverse, not entry-by-entry; that is exactly why the shared determinant

1 - ν_{12} ν_{21}

appears in every

Q

term.

SE5. "

ν_{12}

and

ν_{21}

are independent material constants I must measure separately."

Wrong — they are linked by the symmetry relation

ν_{21} E_{1} = ν_{12} E_{2}

, so only one is independent; measuring both is a redundancy check, not two data points.

SE6. "A quasi-isotropic laminate is isotropic, so I can use isotropic hand-book formulas everywhere."

Wrong — "quasi-isotropic" means only the in-plane extensional stiffness is direction-independent; bending, interlaminar shear, and through-thickness behaviour remain strongly anisotropic.

SE7. "I used the parallel Rule of Mixtures for shear modulus

G_{12}

Wrong — in-plane shear (

G_{12}

, resistance to sliding) passes load through the matrix like the transverse case, so it follows the inverse (series) rule, matching

E_{2}

's form, not

E_{1}

's.

Why questions

WHY1. Why does the matrix, despite being weak, matter at all?

Because it transfers load between fibers, holds their spacing/alignment, and blunts crack tips — without it a bare fiber bundle would buckle in compression and unravel under any off-axis load.

WHY2. Why do we invert compliance to get stiffness rather than working with compliance directly?

Because laminate assembly and FEA (see 5.3-finite-element-analysis) add up stiffnesses of stacked plies to build the laminate response; you cannot simply add compliances of layers that share the same strain.

WHY3. Why does low thermal expansion make composites prized for optics and antennas?

Because near-zero (even negative) fiber expansion can be tuned to hold mirror or antenna geometry stable across the orbital hot/cold swings handled in 4.2-thermal-design, where a metal would defocus.

WHY4. Why does specific stiffness (

E / ρ

), not raw stiffness, drive material selection for spacecraft?

Because every kilogram costs launch mass (2.4-mass-budgets); a composite at 60% of aluminium's stiffness but 60% of its density still wins on stiffness-per-kilogram, which is what sets structural frequency and mass.

WHY5. Why must ply orientation be tracked, not just ply count?

Because each ply only resists load along its fibers; four 0° plies give huge axial stiffness but almost nothing transversely, whereas a balanced 0/±45/90 stack spreads capability where the loads actually point.

WHY6. Why is the term

1 - ν_{12} ν_{21}

always positive and less than one for a real material?

Because thermodynamic stability requires the strain energy to be positive, which forces

ν_{12} ν_{21} < 1

; a value

\geq 1

would imply the material releases energy when loaded — physically impossible.

WHY7. Why compare a composite to aluminium rather than to steel in trade studies?

Because aluminium is the incumbent aerospace baseline; the meaningful engineering question during material selection is "does the composite beat the metal we'd otherwise fly?"

WHY8. Why do sandwich structures pair thin composite faces with a light core?

Because separating the stiff faces (see 3.6.16-sandwich-structures) multiplies bending stiffness for almost no mass, letting a strong-but-thin laminate act like a thick beam.

Edge cases

EC1. What is

E_{2}

in the limit

V_{f} \to 0

(all matrix)?

The inverse rule collapses to

1/ E_{2} = 1/ E_{m}

, so

E_{2} \to E_{m}

— with no fibers the composite is just the matrix, exactly as it must be.

EC2. What happens to

E_{1}

V_{f} \to 1

(theoretical all-fiber)?

E_{1} \to E_{f}

; the parallel rule smoothly recovers the pure-fiber modulus, confirming the formula's end behaviour even though this

V_{f}

is unmanufacturable.

EC3. At

θ = 0°

, what does the transformation matrix

[T]

reduce to?

The identity matrix, so

[\overset{ˉ}{Q}] = [Q]

— a ply aligned with the loading axis needs no transformation, which is the sanity check every rotation formula must pass.

EC4. At

θ = 90°

, how do the roles of

E_{1}

and

E_{2}

swap?

The fibers now run across the load, so the effective axial stiffness becomes

E_{2}

(matrix-dominated) — pulling a "90° ply" is really loading it transversely.

EC5. If a laminate is loaded exactly along one fiber direction but has ±45° plies too, do those angled plies carry zero load?

No — because all plies share the same in-plane strain, the ±45° plies still stretch and carry stress (and shear), just less efficiently than the aligned plies.

EC6. What if

E_{f} = E_{m}

(fiber and matrix equally stiff)?

Both rules give

E_{1} = E_{2} = E_{f}

; with no stiffness contrast the material becomes isotropic and the whole anisotropy story vanishes — the contrast is the composite.

EC7. Is there any orientation where an off-axis ply feels no extension–shear coupling?

Yes — only at

θ = 0°

and

θ = 90°

; at those angles

sin θ cos θ = 0

, so the coupling terms of

[\overset{ˉ}{Q}]

vanish and pure tension gives pure extension.

EC8. What does

ν_{21}

approach as

E_{2} \to 0

(imaginary infinitely soft matrix)?

ν_{21} = ν_{12} E_{2} / E_{1} \to 0

; a vanishingly stiff transverse direction cannot pull the stiff direction inward, so the minor Poisson coupling disappears.