A shop writes in about inconsistent discs. Same SKU, same grit, same carton. On one machine they cut clean. On another they smear, wander, and throw at the rim. The discs are new. The batch is new. The complaint lands on the abrasive.
The pad on that second machine has been on it since spring.
Nobody replaced it, because nothing about it looks broken. A pad does not crack, glaze, or wear visibly. It stops being the surface the disc was designed to sit on, and it does that slowly enough that every disc fitted to it gets blamed in turn.
THE STACK
Four things sit between the motor and the work. The tool, with a speed on its plate. The pad, on the spindle. The disc, on the pad. The work.
Three of them arrive with a specification. The tool has a rated speed and a thread. The disc has a diameter, a grit, a hole pattern and a rated speed. The work has a material and a tolerance.
The pad arrives with whatever was in the box.
On a purchase order the pad has one job, which is to exist. Hardness, face condition and time in service are not fields on a quotation, so nobody measures them. And the pad is the part that decides where the disc sits, whether the disc's face is a plane, and how much of the load reaches the face instead of the rim.
SEVEN EIGHTHS OF THE WORK
A disc on a pad is not one surface at one speed. Speed is proportional to radius, so every ring of the face runs at a different speed, and the spread is not small.
At the rim you are running at twice the speed you are running at half the radius. Averaged across the whole face, the speed is two thirds of the rim. And because area grows with the square of the radius, the outer half of the face — three quarters of the area — carries seven eighths of the sliding work.
That is a wear statement, not a curiosity. Three quarters of the pad's surface is doing seven eighths of the work, and it is also the region furthest from the flange, where any error in the pad's face shows up largest. Whatever the pad does to the disc, it does most of it out there.
THE HALF MILLIMETRE
Pads do not stay flat. They dish. They crown. Dust loads one side before the other. The face the disc sits on stops being parallel to the flange it is clamped against.
Put a number on it, as arithmetic rather than as a measurement, because I have not measured it. A pad half a millimetre out of flat across a 125 mm face. Half a millimetre is 500 micrometres. At P400 the mean grain is 35 micrometres, so the dish is fourteen grains deep at the edge before the disc touches the work. At P240, where the mean grain is 58.5 micrometres, the same dish is eight and a half grains.
A disc cannot be flatter than the surface it is clamped to. A new disc on a dished pad is not new in the sense that matters. It is a disc running on a tilted plane, and the part of it that does the work — that outer seventy-five per cent — is the part furthest out of true.
THE FAILURE WITH NO APPEARANCE
The hook face fails first, and it fails invisibly.
Dust fills the hooks. The loops on the disc back flatten against a face that no longer presents open hooks. Grip falls, and grip is not a cosmetic property. It is the only thing transferring torque from pad to disc. A disc that has lost grip creeps. It walks off centre, and once it is off centre nothing about the disc is balanced any more, and the heavy side is the rim.
The symptoms arrive in this order. The disc wanders and gets re-seated. The scratch smears on one side. Then a disc lets go, and the operator says the batch is bad.
None of that is caused by the disc. All of it is read off the disc, because the disc is the part that gets looked at.
THE NUMBER NOBODY READS
Most pads carry a maximum speed, moulded or printed into the face. Often it is the only number on the pad.
It gets checked at purchase, if at all, and then never again — because the pad is not on the consumables list. The disc is. The pad stays on the tool for months, through tool changes, through a heavier machine arriving to do the same job at a higher setting.
A pad rated for one speed, now running on a tool that goes faster, is not a worn part. It is the wrong part, and nobody reordered it because nobody knew it was a part.
HARDNESS IS THE VARIABLE WITH NO LABEL
Two pads, same diameter, same hook pattern, one soft and one hard, produce visibly different work with an identical disc.
A soft pad conforms. On a radius that keeps the abrasive in contact and spreads pressure, which is what a blend or a repair wants. On a flat panel it also rounds the edge of the contact patch, so the disc cuts slightly crowned and the corners of the part stay soft.
A hard pad stays a plane. On flat work it cuts true and the scratch is even corner to corner. On a curve it concentrates load at the high point, which shows as a burnished band and, held there, as heat.
The grit changes neither outcome. The pad does — and pad hardness is the field on the order that is almost always blank. In the enquiries we receive, which come from shops that already know what they are doing, I have not yet been asked to quote a pad hardness.
TEN MINUTES, NO LAB
The pad is the only part of the stack an operator can evaluate without an instrument.
A straightedge across the face, twice, at ninety degrees. You are looking for light under the middle or under the edge. A dish or a crown shows immediately, and neither is visible to the eye alone.
Press on a fresh disc, lift the tool, hold the disc and twist. It should resist. If it turns under light hand pressure, the interface is finished, whatever the face looks like.
Mount pad and disc, run at working speed with no load, watch the rim. Any vertical movement there is being transmitted into a brand new abrasive layer, and no disc will fix it.
Ten minutes, once a quarter, and the bad-batch conversation mostly stops happening.
WHAT I AM NOT GOING TO CLAIM
No pad life in hours. Pad life is a function of pressure, dust, surface and how the tool is held. Any single number attached to it would be a number I invented.
The arithmetic above is geometry — speed proportional to radius, area proportional to the square of the radius. It is not a test result, and the half millimetre is a round figure chosen to make the comparison. A pad a quarter of that is still, in grain terms, several grains out of true.
What I will say plainly: the pad is the cheapest part of the stack per hour of use, and it is the one that gets replaced years late. That is not a coincidence. It is what happens to a part that has no specification and no price line.
THE DISCS WE MAKE
We supply the coated side of this. Hook-and-loop discs in 5 inch 8-hole and 6 inch 15-hole, aluminium oxide from G60, zirconia and ceramic to G500, film-backed where the finish has to survive a sequence, mesh where dust is the problem. EN 13743, rated 80 m/s, one box minimum, samples free to three pieces.
Whichever side the pad comes from, ask for its rated speed and its face condition with the same seriousness you apply to the disc's. If a machine has been upgraded since the pad was fitted, that is a question to settle before the next carton.
A part with no specification and no price line does not get replaced until it fails visibly. Pads do not fail visibly. They fail in the finish, and the finish gets blamed on the disc.