Causes of Vibration and Spiral Models in CNC Grinding Products
Vibration chatter and spiral models are two different defects, and they rarely share a root cause. This page walks through the mechanical, thermal and process-side origins we check on our own grinding cells, and when grinding is the wrong process for a part in the first place.
Two defects, two diagnostics
Spiral models come from the wheel path; chatter comes from the loop between wheel, spindle and workpiece.
Separate the marks before touching the machine
A spiral model is a periodic pattern that follows the wheel path. Its pitch matches the feed rate divided by the work speed, so the mark spacing tells you the axis that produced it. Chatter is irregular by comparison. It appears as short-wave roughness with no fixed spacing, and its frequency usually sits far above the work rotation.
The first job is measurement, not adjustment. Mark the part in the direction of table travel, photograph the surface at 50× to 100×, and record the spacing. If the spacing changes when you change the work speed, the cause is in the wheel or the dressing system. Spacing that stays fixed points to the machine structure instead.
Keep the part, the setup sheet and the wheel log together. Engineers who send us parts for review get a faster answer when the grinding parameters travel with the sample. We measure the marks, compare them against the recorded feed and speed, and only then open the machine.
The diamond pen sets the wheel, and the wheel sets the surface
A loose or protruding diamond pen cuts the wheel face unevenly. The wheel then carries that shape into every part it touches, and the result looks like a spiral model even though the machine itself is sound. Check the pen seating first. It should be tight, aligned to the wheel centerline and dressed at a consistent infeed.
Dressing depth matters more than most operators expect. A pass that is too deep loads the wheel and leaves a coarse face; a pass that is too shallow glazes it and the wheel rubs instead of cutting. Both conditions raise the risk of chatter. On our surface grinders we keep dressing infeed within the wheel manufacturer's range and re-dress on a fixed part count, not on a feeling.
Wheel grade is the next variable. A wheel that is too hard for the material will glaze, and a wheel that is too soft breaks down fast and loses form. Balance the wheel before mounting it, and re-balance after any dressing that removes significant material. An unbalanced wheel produces a once-per-revolution vibration that shows up as a regular pattern on the work.
Spindle speed, bearings and the thermal loop
Unstable spindle speed shows up as a twisted surface. The wheel slows under load, speeds up when the load drops, and the work surface records that variation. Measure the actual speed with a tachometer and compare it against the drive parameters. A reading that drifts more than a fraction of a percent deserves attention before you change anything else.
Spindle bearings are the other half of that story. Worn or preload-lost bearings let the wheel move in a small orbit. The amplitude is tiny, often under 2 μm, but grinding transcribes it directly onto the part. Listen during a dry run and check runout with a dial indicator on the wheel arbor.
Heat moves the machine while it works. A spindle that warms up over the first hour shifts its centerline, and the cooling fan that is supposed to stabilize temperature can itself become a vibration source when its bearings wear. Let the machine reach thermal equilibrium before finishing passes, and keep the fan and its ducting clean.
Bed, headstock and jaw panel movement
The bed and table carry every force the grinding zone generates. Loose gibs, worn ways or a table that lifts under load all feed vibration back into the wheel. Check the slide clearance and confirm the table moves without stick-slip at slow feed. A table that jumps in small increments will leave marks spaced at the jump interval.
Workholding gets overlooked because it seems passive. A jaw panel or small plate that shifts during the cycle lets the part move a few micrometres, and the wheel cuts a slightly different depth at each pass. Tighten and re-align the jaw panel, then indicate the part after clamping to confirm it has not moved.
Foundation and leveling belong in the same check. A machine that sits on an uneven floor or a cracked pad will hold a twist that changes with part weight. Level the machine, torque the anchor bolts and re-check after the first week of production. This is a slow fix, but it removes a whole class of problems that no parameter change will solve.
Mark pattern to likely cause
Use the spacing and stability of the mark to narrow the list before adjusting parameters.
| Symptom | Likely cause | First check |
|---|---|---|
| Spiral model, pitch tracks feed rate | Wheel face form or dressing fault | Diamond pen seating and dressing depth |
| Regular pattern, once per wheel rev | Wheel out of balance | Re-balance and re-check mounting |
| Irregular short-wave chatter | Spindle bearing wear or runout | Dial indicator on wheel arbor |
| Twisted surface, pitch drifts | Unstable spindle speed under load | Tachometer reading vs drive parameters |
| Marks change with part weight | Bed twist or foundation movement | Level check and anchor bolt torque |
| Marks appear after warm-up | Thermal growth, fan vibration | Thermal soak and fan bearing check |
Common questions
Can a spiral model be removed by a second finishing pass?
Sometimes, but only if the cause is already gone. A lighter pass with a freshly dressed wheel will reduce the visible depth, yet the same vibration source will reprint the pattern at a smaller amplitude. Fix the source first, then take the finishing pass.
How do we tell chatter from a feed mark?
Measure the spacing. A feed mark follows the programmed feed and work speed, so the pitch stays constant when you adjust the speed. Chatter pitch does not follow the program; it follows a structural or spindle frequency that stays fixed while the feed changes.
Does coolant condition affect surface marks?
It can. Dirty or poorly directed coolant changes the thermal load on the wheel and the part, and uneven cooling lets the wheel face load up. Keep the concentration within the supplier range and aim the nozzle at the contact zone, not at the whole wheel.
When is grinding the wrong process for a part?
Grinding suits hardened steel, tight tolerances and fine finishes. It is a poor fit for deep pockets, thin walls that deflect under wheel pressure, and soft materials that load the wheel. Those parts usually run better on a milled or turned operation.
What tolerance and finish can grinding hold in production?
On our grinding and hard-finishing work we hold ±0.005 mm (±0.0002 in) and reach Ra 0.2–0.8 μm on fine finishes. Every part is inspected before shipment, and inspection reports are available on request.
Can you review a part with visible vibration marks?
Yes. Send the part, the drawing and the grinding parameters. We measure the marks and compare them against the recorded feed and speed, then report back what we find. Uploads stay confidential and we can work under an NDA on request.
Send us the part and the parameters
Share a drawing or a marked sample and we will come back with a quotation and a DFM review within 12 hours.
12-hour quote±0.005 mm tolerance100% inspection
