Open the calculatorOpen the calculator

Off-Axis Strength: Theory Against Test Data

One unidirectional ply, pulled at an angle to its fibers. Six criteria, one set of tests.

Problem

A unidirectional E-glass/epoxy ply carries a uniaxial tensile stress σx. Its fibers sit at an angle θ to x. At what σx does it fail, for θ from 0 to 90°? Answer it with max stress, max strain, Tsai-Hill, Tsai-Wu, Hashin and Puck, then compare with the tests.

gripgripθ1, fibersxσxσx

The test: a flat unidirectional strip with its fibers at θ to its length, held in grips at both ends and pulled in tension along x. Repeat it at several θ.

Inputs

MaterialE-glass/epoxy, unidirectional
E153779 MPa
E217926 MPa
G128963 MPa
ν120.25
S1t / S1c1034 / 1034 MPa
S2t / S2c27.6 / 138 MPa
S1255.2 MPa
Loadσx > 0, σy = τxy = 0

These are the glass/epoxy properties in Jones1, converted from ksi and Msi. S12 is the 8 ksi Jones gives for these tests, not the 6 ksi of his property table. Max strain is the only criterion here that needs the elastic constants, because it turns the stresses into strains first. They are also listed so you can rebuild the ply in a calculator.

The stress turns into the fiber axes with Jones Eq. 2.117:

σ1=σxcos⁡2θ,σ2=σxsin⁡2θ,τ12=−σxsin⁡θcos⁡θ\sigma_1 = \sigma_x\cos^2\theta,\quad \sigma_2 = \sigma_x\sin^2\theta,\quad \tau_{12} = -\sigma_x\sin\theta\cos\theta

Every criterion here squares τ12 or takes its magnitude, so its sign does not change the answer. Each criterion then gives the σx that brings the ply to failure.

Strength against fiber angle

The strength axis is logarithmic, because it falls by a factor of almost 40 from 0 to 90°. The chips under the chart hide or show each curve. Move the slider to see the ply stresses at failure and which one each criterion says governs.

015304560759020501002005001000Fiber angle θ (°)σx at failure (MPa)
Max stress
Max strain
Tsai-Hill
Tsai-Wu
Hashin
Puck
Test, read from Jones Figs. 2-37 and 2-401

At θ = 15°, ply stresses when the ply fails, in MPa:

Criterionσxσ1σ2τ12Governing mode
Max stress22120614.8-55.2Shear
Max strain22120614.8-55.2Shear
Tsai-Hill19217912.8-47.9Shear
Tsai-Wu17516311.7-43.7Shear
Hashin19418113.0-48.6Matrix tension
Puck18817612.6-47.1Matrix tension (IFF-A)

In words: all six curves start at 1034 MPa and end at 27.6 MPa. Max stress and max strain sit highest. They share one curve up to about 27°, and past that max strain is higher still. Hashin and Puck run close to Tsai-Hill from 15° on. Tsai-Wu sits lowest and closest to the test points from 5 to 35°. The test points start at 1029 MPa and end at 28.1 MPa.

The test points are Tsai's tension tests, read off Jones Figs. 2-37 and 2-401, which plot the same data. Each reading is good to about ±2% in strength and ±1° in angle. The two figures agree within 4.2% at nine of the ten angles. At 15° they differ by 14%, and the point shown is the mean of both readings.

Results

Largest gap, Tsai-Hill vs max stress
θ = 26.5°
Tsai-Hill 29% lower
Strength lost from 0 to 15°
85% in the tests
81% by Tsai-Hill

Turn the fibers 15° off the load and the ply keeps less than a sixth of its strength. Max stress overpredicts the tests by up to 55% between 5 and 30°, because it checks shear and transverse stress one at a time. Max strain does the same with strains and overpredicts by up to 67%. The two quadratic criteria add the stresses up and land much closer: up to 24% over for Tsai-Hill and 14% for Tsai-Wu. Hashin reaches 24% and Puck 35%.

Uniaxial strength σx at failure, in MPa, and the stress that governs.

θMax stressMax strainTsai-HillTsai-WuHashinPuckGoverns
0°103410341034103410341034Fiber tension
15°221221192175194188Shear
30°11012783.376.683.481.3Matrix tension
45°55.260.249.347.049.348.7Matrix tension
60°36.837.835.334.635.335.1Matrix tension
90°27.627.627.627.627.627.6Matrix tension

At 0° the fibers carry everything and all six give S1t. At 90° the matrix carries everything and all six give S2t. In between, shear governs up to about 27° (where tan θ = S2t/S12) and transverse tension beyond. The Governs column is Tsai-Hill's. The slider readout above gives each criterion's own mode. These numbers match an independent implementation to six significant figures, with its Puck inclination parameters and max strain allowables set to the values used here.

What it means

Small misalignments cost a lot. At 4.6° the tests show 484 MPa, less than half the 0° strength. Fiber waviness and layup error of a few degrees are enough to matter in a unidirectional part.

Max stress compares each stress with its own strength and ignores the rest. At 26.5° the ply carries shear and transverse tension of similar size relative to their strengths, and max stress counts only the larger one. That is where it overpredicts most. Tsai-Hill and Tsai-Wu add the ratios in one quadratic, which is why they follow the tests. Tsai-Wu sits a little lower still because its linear term F2σ2 is positive under transverse tension. At 15.5° both quadratic criteria are still above the test point, by 10% for Tsai-Wu and 20% for Tsai-Hill.

Up to about 27° shear governs both max stress and max strain, and γ12 over its limit S12/G12 is the same ratio as τ12/S12, so the two agree. Past that, max strain sits above max stress. Under σx the fiber stress σ1 pulls the ply in sideways through Poisson's ratio, which cuts ε2. The transverse strain check therefore reaches its limit later than the stress check does. The tests do not show that gain: at 31° max strain is 67% above the test point.

Hashin and Puck check fiber failure and matrix failure separately. Under transverse tension Hashin's matrix mode is (σ2/S2t)² + (τ12/S12)², which is Tsai-Hill without its σ1 terms, so from 15° on the two curves stay within 1% of each other. Puck's matrix check leaves σ1 out until σ1 passes half the fiber strength, and then weakens the matrix only gradually. At 4.6° that puts Puck 35% above the test point, more than either quadratic criterion. From 15° on it sits between Tsai-Hill and Tsai-Wu. At every test point from 5 to 35°, both Hashin and Puck sit further above the tests than Tsai-Wu.

This page plots tension only. Figs. 2-37 and 2-40 also plot Tsai's compression tests on the same material. In compression σ2 is measured against S2c, five times S2t here, and the Tsai-Wu linear term F2σ2 turns negative, so the curves change shape. To check a compressive case, open the calculator below and enter a negative σ1 and σ2.

Related docs:

References

  1. Jones, R.M., Mechanics of Composite Materials, 2nd Ed., Taylor & Francis, 1999. Eq. 2.117, Table 2-3, Figs. 2-37 and 2-40.
Open in calculator

Check one angle yourself. The failure criteria calculator opens with the ply stresses at Tsai-Hill failure for θ = 15° and these strengths.

Open the 15° ply
Written by Rick Schrijver
Share