The Eccentric CNC Follow Up Rusher: How Precision Comes From Following a Lobe

• 1
  What the eccentric CNC follow up rusher actually does
• 2
  Why the follow up device, not the grinder, sets the tolerance
• 3
  When a follow up grinder is the right call, and when it is not
• 4
  What dimensions to specify on an eccentric CNC follow up drawing
• 5
  Five checks that keep an eccentric profile in tolerance
• 6
  Follow up grinding against 5-axis milling for eccentric features

An eccentric CNC follow up rusher is a grinding machine built around one problem: cutting a rotating part whose centerline moves. A camshaft lobe, an eccentric shaft journal or a crankpin is not a cylinder. Its surface sits at a changing distance from the axis, so a fixed wheel position cannot hold size. The follow up device solves that by reading a master cam or a servo command and pushing the wheel in and out in step with the rotation.

The name is a literal description. "Eccentric" is the workpiece geometry. "CNC" is the axis control that moves the wheel head and the work head. "Follow up" is the tracer or servo loop that keeps the contact point on the profile. "Rusher" is shop English for the older mechanical tracer grinders that ran the same job before digital control replaced the master cam.

On a modern machine the master cam is often gone. A rotary encoder on the work spindle feeds position to the CNC, and the control calculates the radial offset for every degree of rotation. The wheel head follows that offset at a bandwidth high enough to hold the lobe through the base circle, the ramp and the nose.

The distinction matters for quoting. A standard cylindrical grinder with a fixed wheel head cannot cut a true lobe. It can only cut a circle. If your drawing shows a lift curve or an eccentricity value, the process is a follow up operation, not plain OD grinding.

• 1
  Eccentric partAny rotating part whose ground surface is off-center: cam, eccentric shaft, crankpin, eccentric bushing.
• 2
  Follow up deviceMechanical tracer or servo loop that moves the wheel in step with spindle rotation.
• 3
  RusherTraditional term for the tracer-type grinder that this process grew out of.

Two error sources dominate on this process. The first is spindle synchronization. The control needs to know where the lobe is at every instant. If encoder resolution is coarse or the loop lags, the nose of the cam gets rounded off and the lift curve falls short of the drawing. Typical closed-loop machines resolve the work spindle to a few arc-seconds, which is enough to hold a cam profile within a few micrometres.

The second is wheel wear. A grinding wheel loses diameter as it cuts. On a fixed-axis grinder that shows up as slow size drift. On a follow up machine it shows up as profile drift, because the wheel radius is part of the offset calculation. The control compensates by tracking wheel dressing and updating the radius value, but the compensation only works if the wheel is dressed on a known schedule.

Machine stiffness sets the ceiling. The wheel head has to accelerate in and out several times per revolution. At 60 rpm that is once per second. At higher speeds the required acceleration rises with the square of the frequency. A light, rigid wheel head with linear motors can follow faster than a heavy hydraulic slide, which is why the same part can hold tighter limits on a newer machine.

In practice a follow up grinder holds ±0.005 mm on diameter and Ra 0.2–0.8 μm on the lobe surface when the wheel is dressed and the coolant is clean. Push the cycle time down or skip a dress and the profile opens up before the diameter does. That is the signature failure of this process.

• 1
  Encoder resolutionSets how accurately the control knows lobe angular position.
• 2
  Wheel radius compensationMust track dressing; a stale radius value shifts the whole profile.
• 3
  Wheel head dynamicsAcceleration limit caps how fast a lobe can be followed without rounding the nose.

Choose the follow up process when the part needs a hardened, ground eccentric surface. Camshafts, eccentric shafts, crankpins and eccentric bushings usually arrive at grinding after heat treatment at 58–62 HRC. At that hardness, milling is not practical and turning is not accurate enough. Grinding removes 0.15–0.35 mm of stock and leaves a surface that seals, bears or runs against a follower.

Choose 5-axis milling instead when the eccentric feature is soft, when the part is a prototype, or when the profile is a freeform surface rather than a lift curve. A 5-axis machining center can produce an eccentric journal to ±0.005 mm in aluminium or pre-hardened steel without a dedicated fixture, and it can do it in one setup alongside the rest of the part. For one-off work that is usually cheaper and faster than building a grinding setup.

The awkward middle is a soft eccentric part with a tight profile tolerance and a fine finish requirement. Milling can hold the geometry but may not reach Ra 0.2–0.8 μm without a second operation. In that case the part is often milled close and then ground, which means two setups and two chances to lose concentricity. Designers can avoid this by relaxing the finish on non-bearing surfaces.

There is also a size limit. Follow up grinding is a between-centers operation, so the part needs centers or a fixture that provides an equivalent axis. Very long, slender eccentric shafts deflect under wheel pressure and need steady rests. Below about 6 mm shaft diameter the deflection usually makes the process impractical.

• 1
  Hardened eccentric surfaces58–62 HRC after heat treatment; grinding is the only practical finish.
• 2
  Soft prototypes5-axis milling holds the geometry without a dedicated grinding setup.
• 3
  Slender shaftsBelow roughly 6 mm diameter, deflection makes follow up grinding unreliable.

Engineers new to the process often dimension the lobe as a radius from the main axis. That is hard to inspect and hard to grind. The working dimensions are the base circle diameter, the cam lift, the lift angle and the nose radius. A cam drawing built on those four values can be checked on a camshaft measuring system and matched directly against the machine offset table.

Eccentricity should be given as a value with a tolerance, not as a nominal center distance alone. On an eccentric shaft the critical stack is the main journal diameter, the eccentric journal diameter and the offset between them. Each needs its own limit. If the offset is open, the assembled mechanism will show runout even when both diameters are in tolerance.

Surface finish belongs on the bearing and follower contact surfaces, not on the whole part. Specifying Ra 0.2–0.8 μm across a full eccentric shaft drives cycle time and cost without adding function. Grind the contact band, leave the rest at Ra 1.6–3.2 μm, and note the transition on the drawing.

Finally, state the reference for runout. Total indicated runout between centers is the normal callout, but some drawings reference the mounting taper. Those are different measurements and they do not agree when the centers are worn. Ambiguity here is the most common reason a first article gets rejected on a part that was ground correctly.

• 1
  Base circle and liftGrind and inspect on lift values, not on radius from the main axis.
• 2
  Offset with toleranceGive the eccentric offset a limit; nominal only will not assemble.
• 3
  Finish where it mattersSpecify fine Ra only on the bearing and follower contact bands.
• 4
  State the runout datumBetween centers and on the mounting taper are not the same measurement.

Check the dress schedule first. On a follow up grinder the wheel radius is a live input to the control. Dressing at a fixed interval and updating the radius value in the control keeps the profile stable. Skipping one dress is the fastest way to lose the nose of a cam.

Check spindle synchronization next. A quick test is to grind a master part, measure the lift curve, and compare it to the previous run. If the curve shifts in phase rather than in amplitude, the problem is synchronization, not size. That points to the encoder or the drive, not the wheel.

Check coolant. Eccentric grinding produces intermittent contact, so heat enters the part in pulses. Poor coolant flow shows up as burn marks on the nose and a soft spot under the surface. Filter the coolant and direct it at the contact zone, not at the whole part.

Check the centers and the steady rest. A worn center shifts the axis and the whole profile moves with it. On slender shafts, set the steady rest close to the lobe and confirm it is not loading the part enough to bend it. Then verify runout on the finished part before it leaves the machine.

Check the inspection method. A follow up ground lobe should be measured on a system that rotates the part and records the lift curve, not on a bench micrometer. Diameter checks alone will pass a part with a rounded nose. If the drawing has a lift curve, the report should have one too.

• 1
  Dress on scheduleUpdate the wheel radius in the control after every dress.
• 2
  Phase shift, not size shiftA curve that moves in phase points to encoder or drive problems.
• 3
  Coolant at the contact zoneIntermittent contact needs directed flow to avoid burn on the nose.