Piston-Rod Surfaces, Hard Chromium and Gland Loading in Hydraulic Cylinders

2. What the gland is asked to do

2.1 Three jobs, three parts

The gland does three different things, and a diagnosis that treats them as one will usually replace the wrong part. The rod seal holds pressure. The wiper keeps dirt, water and ice out on the retraction stroke. The guide rings, usually filled polymer bands, keep the rod central so that the other two can do their work.

The distinction matters because a seal is not designed to be a bearing. A rod seal is a compliant lip, energized by an elastomer or by pressure, whose sealing depends on contact stress distributed evenly around the rod. Push the rod off center and that distribution goes with it: the lip is overcompressed on one side, where it wears and extrudes, and underloaded on the other, where it leaks. The guide rings exist to prevent exactly that, and they can do so only while they carry the side load themselves.

2.2 What the seal needs from the rod

Beneath a working rod seal is a film of oil a fraction of a micrometer thick, drawn out on the extension stroke and, ideally, drawn back in on the retraction. The seal's life depends on that film, and therefore on the texture of the surface that carries it. A surface that is too rough has peaks that cut the lip. A surface that is too smooth holds no oil in its valleys, and the seal runs closer to dry contact, with higher friction and a tendency to stick-slip.

The roughness average Ra cannot tell those two failures apart, because two surfaces with the same Ra can have entirely different shapes: one a plateau broken by occasional valleys, the other a field of peaks. The parameters that separate them are the material ratio defined in ISO 21920-2¹⁴ and the linear material ratio curve of ISO 13565-2¹⁵, whose reduced peak height Rpk measures precisely the part of the surface a seal lip meets first. Seal manufacturers' installation data commonly specify a material ratio as well as a roughness ceiling for this reason. A rod finish specified by Ra alone is under-specified.

3. The running surface: hard chromium over steel

Piston rods for mobile hydraulics are conventionally a quenched-and-tempered or induction-hardened steel bar, of the alloy grades EN 10083-3¹² describes, ground and then electroplated with hard chromium and polished. Engineering chromium is specified by ISO 6158⁶ and by ASTM B650⁷. It is hard, commonly quoted in the range of 800 to 1,000 HV by the method of ISO 6507-1⁸, and it is applied in a layer a few tens of micrometers thick.

Two properties of the coating govern everything that follows. The first is that it is thin and hard over a base that is much softer, several times softer in the case of an unhardened quenched-and-tempered bar. The second is that conventional hard chromium is not a continuous barrier. It deposits with a network of fine cracks, a normal feature of the process rather than a defect, and ISO 6158⁶ classifies coatings by their crack structure for that reason. The cracks hold lubricant in the surface, and they are also the path by which water reaches the steel.

Thickness can be measured without damage by the magnetic method of ISO 2178⁹, since the coating is non-magnetic and the steel beneath it is magnetic, and destructively on a cross-section by the microscopical method of ISO 1463¹⁰. Electroplating introduces hydrogen into the steel, and for the harder substrates the post-plating baking treatment of ISO 9588¹¹ is the established protection against delayed hydrogen cracking.

4. A thin hard layer on a softer base

4.1 Where the stress goes

A stone trapped under a wiper, a grit particle carried in on the rod, or a tool dropped against it applies a local load, and the question is where that load does its damage. The Hertz solution for a sphere pressed against a flat elastic half-space answers it. The radius of the contact is

where F is the load, R the radius of the particle and E* the effective modulus of the pair, about 115 GPa for steel on steel. The shear stress beneath the contact is greatest not at the surface but below it, at a depth that for a Poisson's ratio of 0.3 is

where it reaches about 0.31 of the peak contact pressure. The coating's modulus is similar enough to that of steel for this homogeneous solution to locate the peak approximately. It is not a failure prediction, and it is not used as one.

4.2 The load at which the peak leaves the coating

Setting z_max equal to the coating thickness t and solving for the load gives the load at which the most highly stressed material is no longer the coating but the steel beneath it:

The load scales with the cube of the thickness. Twice the coating carries eight times the load before the peak stress leaves it, and half the coating carries an eighth. Nor is the upper end of the curve reassuring: by the time the peak reaches the interface of a 25 µm coating it is about 1.2 GPa, well above the shear yield strength of an unhardened quenched-and-tempered bar. An unhardened substrate therefore yields under a load well below F_c, and the curve is an upper bound on what an unsupported coating can carry, not a rating.

That is the mechanism behind two established practices. Induction hardening of the rod beneath the chromium raises the strength of the substrate towards that of the coating, so the base can carry the stress when the peak arrives there; the depth of the hardened layer is determined by the method of ISO 18203¹³. And coating thickness belongs to a rod's specification rather than to its finishing, because it enters the load the rod can carry at the third power.

4.3 What a broken coating does to a seal

Once the substrate has yielded, the coating above it has nothing to support it and cracks, and a cracked hard coating over a dented base leaves a raised edge. That edge is harder than any seal material by a wide margin, and it passes under the seal lip on every stroke. A new seal installed over it is being cut from its first cycle. This is the mechanism behind the field observation that a reseal over a damaged rod comes back within weeks: the replacement seal did not fail, it was machined.

5. Corrosion is a storage problem

Chromium resists corrosion because it carries a thin passive oxide film, and that film makes it noble relative to the steel beneath it. Where the coating is continuous this is protection. Where it is not — at a crack that reaches the substrate, at a pore, or at damage of the kind section 4 describes — it becomes the opposite. The steel exposed at the root is the anode of a cell whose cathode is the large surrounding area of chromium, and a small anode coupled to a large cathode corrodes quickly and locally. The pit grows beneath the coating rather than through it, undermining it until the chromium lifts at an edge. That edge then does what section 4.3 describes.

The mechanism needs an electrolyte — water, and particularly salt- or fertilizer-laden water — and it needs time. It does not need the machine to be working. An extended rod on a parked machine exposes its full stroke to weather, dew and wash-down, with no oil film and no wiping, for as long as the machine stands. A working cylinder is repeatedly drawn back through its wiper and re-coated with oil. Pitting on a rod is therefore more often a record of how the machine was stored than of how it was used, and the remedies are correspondingly cheap: retract where the design allows it, and wipe rather than jet-wash, since a pressure jet can drive water and grit past a wiper that was designed to exclude them only on the retraction stroke.

The resistance of a coating is compared by the neutral salt spray test of ISO 9227¹⁶ and the damage rated by the protection-rating scheme of ISO 10289¹⁷. Both are comparisons, not forecasts. ISO 9227¹⁶ itself cautions that salt spray results seldom correspond directly to service life in other environments, and a figure in hours says how one coating ranks against another in the cabinet, not how many seasons a rod will last in a quarry.

6. Side load, and why it concentrates at full extension

6.1 The rod as a lever

Treat the extended rod as a beam supported at two points — the guide rings in the gland and the bearing on the piston, a distance s apart — carrying a transverse load F at the rod eye, a distance L outboard of the gland. A force balance gives the two reactions:

The gland reaction is always greater than the load that causes it, and the multiplier grows as L grows and s shrinks. Extension does both at once: every millimeter of stroke adds a millimeter of overhang and removes a millimeter of bearing spacing.

For the illustrative cylinder of Figure 2, with a one-meter stroke, the gland carries 1.09 times the side load with the rod retracted, twice it at mid-stroke, six times it at nine-tenths of the stroke and twelve times it at full extension. The last tenth of the stroke doubles it. This is why many cylinder designs fit a stop tube, a spacer that limits how close the piston can come to the gland: in the same cylinder a 150 mm stop tube holds the full-extension multiplier to 5.4, at the cost of 150 mm of closed length. Whether a given cylinder has one is a design decision and cannot be seen from outside.

6.2 Where the side load comes from

A cylinder pinned at both ends is, in principle, a two-force member: it can carry force only along the line joining its pins, and so carries no side load at all. Three things break that principle in service.

The first is pin friction. A pin joint that resists rotation can transmit a moment, and its line of force is displaced from the pin center by the radius of the friction circle,

for a pin of radius r and friction coefficient μ. For a 50 mm pin, a lubricated coefficient of 0.15 gives about 3.7 mm, and a dry or galled joint at 0.5 gives about 11 mm — about three times as much.

The second is a bent rod and the third a mounting out of line, and both displace the axial force from the axis of the bearings in the same way. Whatever its source, an axial force F_ax acting at an eccentricity e from the bearing axis is reacted by a couple at the gland and piston bearings:

For the cylinder of Figure 2 at full extension, with 100 mm between the bearings and 100 kN of axial force, lubricated pins can put up to about 3.7 kN across the guide rings and dry pins up to about 11 kN. That is the arithmetic behind renewing the pins and bushes whenever a cylinder is off: a replacement installed on the worn joints that loaded its predecessor is loaded the same way.

A bent rod also worsens itself under compression, because the axial force acts on the bend and increases it. ISO/TS 13725⁵ gives the method for determining the buckling load of a cylinder, and that calculation assumes a straight rod.

6.3 What the gland shows

The consequence is visible on disassembly. Guide rings loaded within their capacity wear evenly around the circumference. Guide rings carrying a persistent side load wear on one side, and once worn through they let the rod bear on the metal of the gland. From then on the seal is the bearing, and it wears on one side only: overcompressed and extruded where the rod bears, relaxed and leaking opposite. One-sided wear on the rings and the seal is the signature of side load, and it places the cause outside the gland.

7. Implications for inspection and for replacement

Three consequences follow for practice, and none of them is a specification.

The first is that the rod decides the repair. A seal kit restores the parts of the gland that wear by design; it restores nothing on the rod. A rod that is scored, pitted or lifting at an edge will cut a new seal as it cut the old one, and the useful inspection is of the rod under good light along its full stroke, before any parts are ordered. Where there is doubt about how much coating remains, ISO 2178⁹ measures it without damage.

The second is that the wear pattern names the cause. Even wear on the guide rings and the seal is service; one-sided wear is side load, and side load comes from the pins, the straightness of the rod or the mounting — none of which a new seal or a new cylinder changes. Renewing the pins and bushes and checking the rod for straightness belong to the same job.

The third is that corrosion found on a rod is evidence about storage, and the remedy is operational rather than metallurgical. A rod with a thicker or crack-free coating will last longer under the same neglect. It will not be immune to it, because the mechanism of section 5 needs only an interruption in the coating, and time.

The general principle is the one that runs through this library. A replacement seal inherits the rod, and a replacement cylinder inherits the pins, the mounting and the way the machine is parked. Establishing which of those was the cause is what decides whether the repair lasts.