Evaluating a hydraulic hose involves two entirely different questions: how much pressure it can take, and how many pressure cycles it can take. The first is strength, answered by burst testing. The second is endurance, answered by impulse testing. They do not always move together — a hose with a very high burst pressure can have a modest impulse life, which is why "the expensive hose is not always the durable one" on high-cycling duty.
The standard method behind the second question is ISO 6803, Rubber or plastics hoses and hose assemblies — Hydraulic-pressure impulse test without flexing (current edition 2017; adopted domestically as GB/T 5568).
The rig is not complicated in principle: fit the assembly with couplings, fill it with hydraulic fluid held at a specified temperature, then pressurise and release repeatedly until the specimen fails or reaches the required number of cycles. Every parameter, however, is tightly defined, because any freedom destroys comparability.
Specimens and geometry. A group is normally four unaged assemblies. Free length follows the relationship given in the standard (L = 2d + 2r), and the bend configuration depends on bore: specimens of 22 mm ID and below are bent through 180°, larger bores through 90°. What is tested is never a straight length of hose but an assembly in a bent state — close to how it is actually installed.
Fluid and temperature. The test fluid is typically an ISO VG 32 to VG 100 hydraulic oil, circulated to keep temperature uniform. Test temperature is set by the referring hose standard; ISO 6803 lists preferred values of 85, 100, 120, 125, 135 and 150 °C, usually with ±3 °C tolerance. Temperature is pushed high because heat both softens the compound and accelerates ageing — one of the most effective ways to shorten test duration.
Waveform and control. This is the part most easily overlooked and most capable of skewing results. The standard defines a waveform envelope and requires the rate of pressure rise to follow R = f(10p − k) with k = 5 MPa, allowing roughly ±10% tolerance. The measurement system is specified too: the pressure recorder must have a natural frequency above 250 Hz with critically damped response. The reason is practical — if you cannot measure the true peak, the test means little. High-pressure impulse frequency typically falls in the 0.5–1.3 Hz band.
Thresholds and criteria. Passing means reaching the specified cycle count without leakage or burst. Note that leakage below class 4 of ISO/TR 11340 is not counted as failure — weeping is tolerated, leaking is not. Cycle thresholds differ greatly by product standard: commonly compiled figures put single- and double-wire braid constructions in the 150,000–200,000 range and four-ply spiral constructions in the 400,000–500,000 range, with test pressures usually at 120–133% of rated working pressure and different oil temperatures for braid versus spiral (spiral types often at 120 °C). Compilations vary slightly between sources; the product standard text governs.
This is an accelerated test, not a calendar. That point cannot be over-emphasised. Over-pressure, elevated temperature and a fixed waveform compress years of service into weeks on the bench, but real duty adds temperature cycling, a random load spectrum, installation stress, fluid ageing and environmental attack. More importantly, over-acceleration changes the failure mode — what fails on the rig may not be what fails in the field. Impulse count is therefore best used as a comparative yardstick: a way to rank constructions and suppliers under one method, not as a direct prediction of field life.
That is also why a practice has emerged of calibrating bench results against field data, using real duty feedback to correct laboratory conclusions. For a manufacturer, the genuine accumulation of know-how is not merely "our samples reached one million cycles" but "what one million cycles of ours corresponds to on the customer's machine."