Stress concentration factor calculator

Get Kt and the peak stress at a hole in a plate: a circular hole in a plate of any width under tension, an elliptical hole, or a hole under biaxial stress or shear. Enter the peak stress from your ANSYS, HyperMesh or SOLIDWORKS model and the calculator tells you whether it agrees. Free, no sign-up, runs in your browser.

Pilkey / Peterson curve fit · Inglis · Kirsch · every value checked against Skill-Lync's own finite-element model · Updated 19 Nov 2026

Tension along the plate's length, hole in the middle of the width, plate long enough that the load is uniform a hole-width or so away. Linear elastic; Kt does not depend on the material. Net-section Kt from Pilkey's fit of Howland's solution, checked against our FE model.

Kt (net section)2.156gross-section Kt 4.312
Peak stress1.20 MPaσnet 0.556 MPa · σgross 0.278 MPa
σx (MPa) against distance from the hole edge (mm) 00.510102030σnet 0.556 MPaσmax 1.21 MPa (FE) distance from the hole edge (mm) σx (MPa)

    What is a stress concentration factor?

    Cut a hole in a plate and pull it. The load that used to flow straight past has to go round the hole, and the lines of force crowd together at its sides, like traffic squeezing past a lane closure. The stress at the edge of the hole, on the line across the load, is several times the stress in the rest of the plate. The stress concentration factor Kt is that ratio:

    For a small circular hole in a wide plate Kt = 3, whatever the material or the size of the hole. That number comes from Kirsch's elasticity solution of 1898. The peak sits at the two points of the hole edge on the line across the load. At the two points on the load line the hoop stress is −σ, a compression, which surprises most people the first time they plot it.

    The 300 × 120 mm plate with a 60 mm hole, pulled at both ends; the peak stress is at the top and bottom of the hole peak
    The plate from the worked example below, to scale: pulled at both ends, peak stress at the top and bottom of the hole.

    Kt is a property of the shape only, for a linear elastic material. It tells you the real peak stress in a part whose nominal stress you can work out by hand, and it is the number you compare a finite-element result against.

    Kt for a plate of finite width

    In a real plate the hole takes up some of the width, and Kt depends on d/W. Howland solved this in 1930. The curve fit printed by Pilkey and in Peterson's Stress Concentration Factors (chart 4.1) is based on the net section, the material left beside the hole:

    Skill-Lync checked it with its own finite-element model: a plane-stress quarter plate with 9-node elements, refined until Kt stopped changing (in the fourth decimal), with a plate three widths long so the load is uniform. The fit is within 1.0% of the FE everywhere from d/W = 0.05 to 0.7, and slightly low, so it errs a little on the unsafe side for large holes:

    d / WKtn, formulaKtn, Skill-Lync FEFormula vs FEKtg (gross), FE
    0.052.8522.858−0.21%3.008
    0.12.7212.732−0.39%3.035
    0.22.5062.522−0.60%3.152
    0.32.3472.362−0.64%3.374
    0.42.2332.248−0.68%3.747
    0.52.1562.174−0.83%4.348
    0.62.1062.128−1.01%5.319
    0.72.0752.097−1.03%6.989

    Read the two columns together. Net-section Kt falls as the hole grows, from 3 towards 2, because the remaining ligament gets narrower and carries the load more evenly. Gross-section Kt rises, to 6.99 at d/W = 0.7, because the nominal stress it divides by ignores the hole. Both give the same peak stress. Mixing them up, a net-section Kt times a gross-section stress, under-predicts the peak by a factor 1 − d/W: half, for the plate below.

    Worked example: the 300 × 120 mm plate with a 60 mm hole

    The classic first FEA assignment (Skill-Lync's FEA courses set it, and many student reports answer it): a steel plate 300 mm long, 120 mm wide and 30 mm thick with a 60 mm hole in the middle, fixed at one end and pulled with 1,000 N at the other. d/W = 0.5:

    Our converged FE model of this plate gives 1.208 MPa, 0.9% above the formula. The stress across the net section, from the edge of the hole to the edge of the plate, starts at 1.208 MPa, falls below the average of 0.556 MPa 42% of the way across and is down to 0.195 MPa at the plate edge:

    σx (MPa) against distance from the hole edge (mm) 00.510102030σnet 0.556 MPaσmax 1.21 MPa (FE) distance from the hole edge (mm) σx (MPa)
    σx across the net section of the 300 × 120 × 30 mm plate under 1,000 N (Skill-Lync FE). The dashed line is the average, σnet.

    Two cautions for the assignment. The stresses are tiny, about 1 MPa, because 1,000 N is a small load for a 30 mm steel plate: that is fine for studying Kt, which does not depend on the load. And the fixed end matters near the clamp, not at the hole: a fully fixed face stops the plate narrowing (Poisson's effect), which puts small stress peaks at the corners of the fixed face. Those are an artefact of the support, so read the peak at the hole, 150 mm away.

    How to check an FEA result for a plate with a hole

    Model the plate, read the peak stress at the hole, and compare it with Kt × σnet. If it is more than a few percent off, the mesh is usually the reason. We meshed the same plate with linear 4-node and quadratic 9-node quadrilaterals of increasing fineness and read the nodal (averaged) stress at the hole, as a post-processor shows it by default:

    Elements per quarter holeElement lengthLinear (4-node)Quadratic (8/9-node)
    223.56 mm0.906 MPa (−25.0%)1.090 MPa (−9.8%)
    411.78 mm1.063 MPa (−12.0%)1.189 MPa (−1.6%)
    85.89 mm1.160 MPa (−4.0%)1.207 MPa (−0.1%)
    162.95 mm1.197 MPa (−0.9%)1.208 MPa (+0.00%)
    321.47 mm1.206 MPa (−0.1%)not run
    Converged1.208 MPa

    A coarse mesh always reads low here, because the elements average the stress over a region where it falls steeply. With linear elements you need about 16 elements on a quarter of the hole, an element length near d/20, to get within about 1%. Quadratic elements get there with 8, about d/10. Check convergence the same way on your own model: halve the element size at the hole and see if the peak moves.

    • Read the right stress. At the hole edge the material is in plain uniaxial (hoop) stress, so von Mises, maximum principal and σx are all equal at the peak point. Elsewhere they differ.
    • Nodal, not element stress. Element (centroid or unaveraged) values sit inside the element and read lower than the edge.
    • Shell or 2D vs solid. For a thin plate, a shell or plane-stress model and a solid model agree. A thick solid plate has a slightly higher peak at mid-thickness than at the faces.
    • Higher than the formula? Look for a peak at the supports or a point load, and make sure the load is spread over the end face.

    Elliptical holes, slots and cracks

    Inglis extended the solution to an elliptical hole in 1913. With a the half-axis across the load and b the half-axis along it:

    ρ is the radius at the sharp end. A circle (a = b) gives 3 again. Turn a 2:1 ellipse so its long axis lies across the load and Kt = 5; along the load it is 2. That is why slots are cut along the load path, and why a crack across the load is so dangerous: as b goes to zero the tip radius vanishes and Kt grows without limit. Fracture mechanics replaces Kt with the stress intensity factor for exactly that reason. Our FE model agrees with Inglis to within 0.4% for a/b from 0.5 to 3 (in a plate 80 half-axes wide, which adds a little):

    a / b0.5123
    Kt, Inglis2.0003.0005.0007.000
    Kt, Skill-Lync FE2.0013.0055.0087.028

    Biaxial stress and shear: Kirsch's solution

    When the plate is loaded both ways, the hoop stress round the hole edge is the sum of the two uniaxial patterns, with θ measured from the σ1 direction:

    Loadingσ2 / σ1σθ at 90°, KirschFEσθ at 0°, KirschFE
    Pure shear-14.00 σ14.009 σ1-4.00 σ1-4.009 σ1
    Tension + compression-0.53.50 σ13.507 σ1-2.50 σ1-2.506 σ1
    Uniaxial03.00 σ13.005 σ1-1.00 σ1-1.004 σ1
    Thin cylinder skin0.52.50 σ12.503 σ10.50 σ10.498 σ1
    Equal biaxial12.00 σ12.001 σ12.00 σ12.001 σ1

    Equal tension both ways gives 2σ all round the hole, lower than uniaxial tension, because σ2 adds −σ2 at the points where σ1 peaks. Shear is the worst case at ±4τ, which is why holes in torque tubes and shear webs get doublers or flanged edges. The skin of a thin pressurised cylinder, with hoop stress twice the axial stress, gives 2.5 times the hoop stress at the hole, before the effect of the shell's curvature, which raises it further for holes that are large compared with √(Rt).

    How far does the stress concentration reach?

    Not far. Along the line across the load, in a wide plate, Kirsch's solution gives

    with r the hole radius and y the distance from the hole centre. One radius out from the edge (y = 2r) the stress is down to 1.22 σ, and at y = 3r it is 1.07 σ. Our FE model matches this curve to within 0.2% (in a plate 80 radii wide). So the raised stress is a local effect, about one hole radius deep, which is why it matters most for fatigue and brittle fracture and least for the overall strength of a ductile part.

    When to use Kt in design (and Kf)

    • Brittle materials (cast iron, glass, some composites and ceramics): apply Kt. The material cannot yield locally to relieve the peak.
    • Ductile metals under static load: design texts such as Shigley's usually do not apply Kt, because a small region yields and the stress redistributes. If Kt × σnet passes yield, a linear FE model's peak is no longer the real stress; run a plastic analysis if it matters.
    • Fatigue: use the fatigue notch factor Kf = 1 + q(Kt − 1), where the notch sensitivity q (0 to 1) depends on the material's strength and the notch radius; read it from Peterson's or Neuber's curves for your steel or aluminium. Large holes in high-strength steels have q close to 1, so Kf ≈ Kt.

    What this calculator does not cover

    • Other shapes: shoulder fillets, grooves and notches in bars and shafts, keyways, threads. Peterson's handbook has charts for all of them.
    • Several holes, holes near an edge, pin-loaded holes (lugs), and bending of the plate. Holes interact when they are closer than a few diameters; model them in FE or use Peterson's charts.
    • Plasticity, fatigue life and cracks. Kt is elastic. For a crack use the stress intensity factor, not Kt.
    • Short plates. The values assume the load is uniform some way from the hole, roughly one plate width.
    Learn FEA on real parts

    Skill-Lync's FEA courses start with exactly this plate and go on to meshing, convergence and structural analysis of real components in Ansys and HyperMesh, the work CAE teams hire for.

    Frequently asked questions

    What is the stress concentration factor of a plate with a hole?

    For a small circular hole in a wide plate under tension, Kt = 3: the stress at the edge of the hole, at the two points on the line across the load, is three times the stress far away. In a plate of finite width it is usually quoted on the net section (load divided by (W − d) × t), and falls from 3 towards 2 as the hole gets bigger: 2.51 at d/W = 0.2 and 2.16 at d/W = 0.5.

    Is Kt based on the gross or the net area?

    Either, but you must use the stress the chart was made with. Peterson's chart and the formula on this page give Ktn, based on the net section. Ktg, based on the gross section, is Ktn / (1 − d/W): at d/W = 0.5 Ktn = 2.16 but Ktg = 4.31. Multiply each by its own nominal stress and you get the same peak stress.

    What is the stress concentration factor for the 300 × 120 mm plate with a 60 mm hole?

    d/W = 0.5, so Ktn = 2.156 from the formula. With 1,000 N on a 30 mm thick plate, σnet = 0.556 MPa and the peak stress is 1.198 MPa. Skill-Lync's converged finite-element model of the same plate gives 1.208 MPa (Ktn 2.174), 0.9% higher than the formula.

    Why is my ANSYS or HyperMesh peak stress lower than the hand calculation?

    Usually the mesh is too coarse at the hole. In our model of the 60 mm hole, 8 linear (4-node) elements on a quarter of the hole edge read 4.0% low and 4 elements 12% low; 16 elements (about 2.95 mm each) bring it within 0.9%. Quadratic elements converge much faster: 8 of them are within 0.1%. Also read nodal (averaged) stress, not element centroid values.

    What is the stress concentration factor of an elliptical hole?

    Kt = 1 + 2a/b in a wide plate (Inglis, 1913), where a is the half-axis across the load and b the half-axis along it. A hole twice as long across the load as along it has Kt = 5; the same ellipse turned the other way has Kt = 2. In terms of the tip radius ρ = b²/a, Kt = 1 + 2√(a/ρ), which grows without limit as the ellipse closes into a crack.

    What is Kt for a hole under biaxial stress or shear?

    For a circular hole in a wide plate, the hoop stress at the edge is σ1(1 − 2cos 2θ) + σ2(1 + 2cos 2θ) (Kirsch). Equal tension both ways gives 2σ everywhere round the hole; pure shear (σ2 = −σ1) gives ±4σ1; a thin cylinder's skin, with hoop stress twice the axial stress, gives 2.5 times the hoop stress, before shell curvature effects.

    What is the difference between Kt and Kf?

    Kt is the theoretical (elastic) factor that depends only on geometry and loading. Kf, the fatigue notch factor, is what a real material shows in a fatigue test: Kf = 1 + q(Kt − 1), where the notch sensitivity q is between 0 and 1 and depends on the material and the notch radius (Peterson's and Neuber's curves). Sharp notches in soft materials have q well below 1; large holes in high-strength steel are close to 1.

    Related on Skill-Lync

    Sources

    • W. D. Pilkey, Formulas for Stress, Strain, and Structural Matrices, 2nd ed., Wiley 2005 (finite-width plate with a central circular hole in tension, net-section curve fit); W. D. Pilkey and D. F. Pilkey, Peterson's Stress Concentration Factors, 3rd ed., Wiley 2008, chart 4.1.
    • R. C. J. Howland, "On the stresses in the neighbourhood of a circular hole in a strip under tension", Philosophical Transactions of the Royal Society A 229 (1930) 49-86.
    • C. E. Inglis, "Stresses in a plate due to the presence of cracks and sharp corners", Transactions of the Institution of Naval Architects 55 (1913) 219-241.
    • G. Kirsch, "Die Theorie der Elastizität und die Bedürfnisse der Festigkeitslehre", Zeitschrift des Vereines deutscher Ingenieure 42 (1898) 797-807; in English in S. P. Timoshenko and J. N. Goodier, Theory of Elasticity, McGraw-Hill.
    • R. G. Budynas and J. K. Nisbett, Shigley's Mechanical Engineering Design, McGraw-Hill (stress concentration under static and fatigue loading, notch sensitivity).
    • Finite-element results: Skill-Lync, scripts/verify/kt_plate.mjs: plane-stress quarter model with symmetry, 9-node and 4-node quadrilaterals graded towards the hole, direct (Cholesky) solve, nodal-averaged stresses; no stress concentration formula is used. This page checks its formulas against those results at build time.
    Free counselling

    Go from hand calcs to FEA you can defend.

    CAE roles want engineers who can mesh a part, prove the mesh has converged and explain the result, not just run the solver. A Skill-Lync counsellor can map the course that fits the role you want.

    Talk to a CAE course counsellor

    Free. We call you back, usually the same day.

    By submitting you agree to our Terms and Privacy Policy.

    Talk to a career counsellor

    Talk to a career counsellor

    Free 20-minute call. We'll map the right course to where you want to work.

    By submitting you agree to our Terms and Privacy Policy.