How a Fully Automatic Oil Films System Holds Precise Power Accuracy Control
A feeder that runs on an oil film does not hold position with clamps. It holds it with pressure, speed and film thickness kept inside a narrow band. This page explains where that band comes from, how the control loop reads it, and which bar sizes and surface conditions fall outside it.
Why fully automatic oil films carry the load instead of a clamp
A fully automatic oil films feeder pushes bar stock into a lathe or grinder through a sleeve whose bore is a few hundredths of a millimeter larger than the bar. The gap fills with oil. While the bar turns, it does not touch the sleeve wall, so the surface is not scratched and the drive motor is not fighting sliding friction.
The film does two jobs at once. It centers the bar, because pressure around the circumference stays roughly equal. It also damps vibration, because oil has to be squeezed out of a narrow gap before the bar can move sideways. That squeeze takes time, and the delay is what turns a thin film into a stiff spring.
Film thickness is not fixed. It settles where the hydrodynamic pressure from rotation balances the push from the feed cylinder. Spin faster and the film thickens. Push harder and it thins. The controller has to keep both inputs inside a window, or the bar drifts off center or starts to rub.
This is why the sleeve bore is matched to the bar, not to the machine. A 20 mm bar in a 20.05 mm sleeve behaves well. The same bar in a 20.20 mm sleeve loses centering force and rattles. In a 20.01 mm sleeve it starves for oil and seizes. The window is narrow by design.
What the control loop actually measures
The loop closes on three signals. A servo encoder reports how far the bar has advanced. A pressure transducer reports the push from the feed cylinder. A proximity or laser sensor watches bar runout near the sleeve exit. Together they describe position, force and alignment.
Speed and force are coupled. If the controller only regulates speed, a harder push from a bent bar raises film pressure until the bar rubs. If it only regulates force, a sticky bar lets speed sag and the lathe starves. Good controllers regulate force as the inner loop and let speed follow as the outer loop.
Resolution matters more than raw speed. A servo that resolves 0.01 mm per pulse can hold feed to ±0.05 mm per revolution without hunting. A coarser encoder forces the loop to gain aggressively, and the bar starts to oscillate at the sleeve exit. The oscillation shows up as a wavy turned surface.
Sampling rate sets the ceiling. Below about 500 Hz, the loop cannot see the once-per-revolution pulse of a bar spinning at 3,000 rpm. The controller then corrects a problem it has already passed, which adds phase lag. Phase lag is the main reason a feeder that runs fine at 1,500 rpm becomes unstable at 4,000 rpm.
Where accuracy control starts to fall apart
Bar straightness is the first limit. Feed accuracy is measured at the sleeve, but the bar has to reach it. A bar with 0.5 mm per meter of bow will push the film to one side no matter how good the controller is. Under 0.2 mm per meter, most loops hold easily.
Surface finish on the bar matters too. Ground and polished stock draws a stable film. Hot-rolled or scaly stock tears the film locally, and the pressure transducer reads spikes instead of a smooth curve. Those spikes are false corrections waiting to happen.
Temperature shifts the oil viscosity. A feeder that is accurate at 20 °C in the morning can drift by mid-afternoon when the hydraulic reservoir reaches 45 °C. Viscosity drops, the film thins, and centering force falls with it. Oil coolers are not optional on long runs.
Bar size sets the practical window. Below about 3 mm, the film is too thin to carry meaningful load and the bar behaves like a wire. Above roughly 80 mm, the required film pressure and oil flow grow faster than the sleeve can supply. Both ends need a different feeding method, not a tuning change.
Bar condition versus expected feed behavior
Use this as a first pass. Values assume ground stock and a matched sleeve.
| Bar condition | Film stability | Typical action |
|---|---|---|
| Ground, bow under 0.2 mm/m | Stable across the speed range | Run as-is, standard tuning |
| Ground, bow 0.2–0.5 mm/m | Stable below 2,000 rpm | Reduce feed force, add a guide bushing |
| Hot-rolled, light scale | Local pressure spikes | Brush or peel the bar first |
| Cold-drawn, ovality over 0.05 mm | Cyclic film loss each turn | Match sleeve to the larger diameter |
| Bar under 3 mm | Film cannot carry load | Switch to a collet or belt feeder |
| Bar over 80 mm | Oil flow becomes the limit | Use a separate pressure lubrication unit |
| Oil above 45 °C | Film thins, centering drops | Add an oil cooler and re-tune |
When the oil film approach is the right call
If your stock is ground, straight and between 3 and 80 mm, an oil film feeder gives the best surface and the longest sleeve life. If the bar is scaly, badly bowed or outside that size range, fix the stock or move to a collet feeder first. Tuning a good controller cannot repair bad bar.
Common questions
Does higher oil pressure always improve centering?
No. Centering force comes from the pressure difference around the bar, not from absolute pressure. Raising supply pressure past the point where the film is fully developed adds heat and flow without adding stiffness.
On most sleeves the useful range is 0.3–0.8 MPa at the inlet. Above that, the controller sees a flatter pressure signal and loses the sensitivity it needs to correct small errors.
How often does the oil need to be replaced?
It depends on contamination, not on hours alone. Fine chips from the turning operation migrate back along the bar and collect in the sleeve. Once the filter differential pressure rises, film thickness becomes unpredictable.
Check the filter on a fixed interval and sample the oil for particle count. Clean oil keeps the loop predictable; dirty oil turns every reading into noise.
Can the same feeder handle both turning and grinding?
Usually yes, with different speed and force setpoints. Grinding runs at higher spindle speed but often lower feed force, so the film runs thinner. Save the two setups as separate recipes rather than re-tuning by hand.
If the grinder removes less material per pass, a lower feed speed is fine. The loop still holds position because the encoder, not the cutting load, defines the reference.
What causes a wavy surface on the turned part?
A wave at a fixed pitch almost always points to loop oscillation, not to a bad tool. The feeder is correcting faster than the bar can respond, and the correction repeats once per revolution.
Reduce loop gain first, then check encoder resolution. If the wave pitch changes with spindle speed, the problem is in the control loop. If it stays fixed, look at the sleeve or the bar.
Is an oil film feeder safe with short bar remnants?
The last 200–300 mm of a bar has less mass and less length inside the sleeve, so centering force drops. Most controllers will keep feeding but the accuracy claim no longer holds.
Set a remnant length in the recipe and stop the feed before that point. A short remnant that whips inside the sleeve can damage both the bar and the bore.
Do we need a different controller for different materials?
Different recipes, usually the same hardware. Aluminium runs with lower feed force and a thinner film; stainless needs more force and generates more heat in the oil.
The pressure and speed setpoints change, but the loop structure does not. Store material-specific recipes and verify the first part of each run.
Send us your bar size and we will check the window
Tell us the bar diameter, material and straightness spec. We will confirm whether an oil film feed holds your tolerance or whether a different method fits better.
12-hour quote100% inspection
More machining notes
We publish setup notes, tooling trials and inspection data from the factory floor.
