Presets
Pressure experiments
Autofrettage presses the bore to the target pressure, holds, and releases. The tube remembers the highest pressure it has seen.
Geometry
Material
Pressures
The effective autofrettage pressure is min(target, p_F, 2 p_y): above 2 p_y the bore would yield again on unloading, which this model does not follow.
Camera
Layers
Quality
Experiment
Model checks
The science
The problem: a thick wall under internal pressure
A long tube with bore radius a and outer radius b carries internal pressure p. Radial equilibrium of a thin ring gives d(sr)/dr + (sr - st)/r = 0 (sr radial, st hoop stress). In the elastic range the solution is the Lame pair sr = A - B/r^2, st = A + B/r^2 with A = p a^2/(b^2 - a^2) and B = A b^2. The hoop stress is highest at the bore, st(a) = p (b^2 + a^2)/(b^2 - a^2), and the Tresca difference there is st - sr = 2 p b^2/(b^2 - a^2). For a thin wall this reduces to Barlow's p r/t (one of the checks). Making the wall thicker does not help beyond a limit: the bore stress can never fall below p, however large b is.
Autofrettage: yield the bore on purpose
The tube is first pressurised so far that the inner part yields, then unloaded. Material is treated as elastic-perfectly-plastic with the Tresca condition |st - sr| <= sy. Yield starts at the bore when p_y = sy (b^2 - a^2)/(2 b^2). A plastic zone a < r < c grows with pressure: p(c) = sy [ ln(c/a) + (b^2 - c^2)/(2 b^2) ], with sr = -p + sy ln(r/a) and st = sr + sy inside it and a Lame solution under the interface pressure pc = sy (b^2 - c^2)/(2 b^2) outside. When c = b the wall is fully plastic, p_F = sy ln(b/a).
Why it helps: residual compression
Unloading is elastic, so the residual stress is the loaded stress minus the elastic Lame stress of the same pressure. The outer, still elastic shell springs back and squeezes the permanently stretched bore: the residual hoop stress at the bore is large and compressive (-313 MPa for the default tube). In service the pressure adds the ordinary Lame stress on top of this, so the bore sees much lower peak tension. The stress range is unchanged (the elastic response is linear), but the mean stress drops, which is what fatigue cares about. A Morrow mean-stress correction sa = (sf - sm)(2N)^b with sf = su + 345 MPa and b = -0.085 turns that into a life estimate. This is an illustration, not a design number; reported life gains for real autofrettaged barrels and fluid ends are generally a factor of a few to tens, and the extrapolated factor here is often larger.
Limits and checks
Reverse yielding (Bauschinger effect) would start when the unloaded bore reaches -sy, i.e. above 2 p_y; this model caps the autofrettage pressure at min(p, p_F, 2 p_y) and says so. Four independent checks run on every change: (1) force balance of a half tube, the integral of the hoop stress over the wall equals p a (the residual state must integrate to zero and have zero radial stress on both surfaces); (2) numerical integration of the radial equilibrium equation with 20 000 steps against the closed form p(c); (3) the thin-wall limit against Barlow; (4) a yield-condition audit over 2001 radii. The yield-onset experiment scans the pressure on a virgin tube and compares first yield and full overstrain with p_y and p_F.
What is simplified
Perfectly plastic material (no strain hardening), Tresca criterion, small strains, an infinitely long tube whose axial stress is not part of the yield test, no thermal or friction effects, and a single fixed material. Real autofrettage uses Mises or hardening models, accounts for the Bauschinger effect and end effects, and verifies the result by strain gauges or X-ray diffraction. Fatigue life is a one-line Morrow estimate. All values are representative of a high-strength steel, not a certified design.
What am I looking at
- Cut face: the front face of the tube is painted with the hoop stress through the wall: blue is compressive, red is tensile, grey is near zero. The dashed green ring marks the edge of the plastic zone
c. - Gauge and hand pump: the dial reads the pressure in the bore (MPa); the lever swings when the pressure changes.
- Wall board: a live plot of hoop stress against radius: current (red) and residual after unloading (purple).
- Graph 2: the plastic radius against pressure. The white dot is the highest pressure the tube has seen, the cyan dot the pressure now.
What to try
- Choose "Plain tube", press Autofrettage: nothing yields because the target pressure is zero. Then pick "Moderate autofrettage" and press it again: the cut face turns blue at the bore after unloading.
- Press "Service cycles" and compare graph 4: the green curve starts below zero at p = 0 because of the residual compression, the dashed red curve is the plain tube.
- Pick "Near full overstrain", then raise the working pressure above the autofrettage pressure: the bore yields again in service (a warning appears) and the benefit is lost.
- "Thin wall" makes p_y and p_F almost equal: the plastic zone races through the thin wall.
- Press "Yield-onset test" and read measured against analytic p_y and p_F.