| What you see | Check these first, in this order |
|---|---|
| Belt slides, won't cut | Contact wheel too soft → tension too low → speed too low |
| Workpiece blues or burns | Belt already dull → speed too high → too much pressure in one pass |
| Evenly spaced marks | Joint impact → contact wheel too hard → tension too high |
| Scratches too deep for the grit | Grit step skipped → previous scratches never removed → contamination |
| Belt wanders off the wheel | Tension too low → wheels not parallel → belt length wrong |
| Belt tears at the joint | Fitted backwards → tension too high → joint hitting an obstruction |
| Belt clogs up quickly | Material too soft/gummy for the grit → no anti-clog coating → speed too low |
| New belt behaves badly from the start | Storage humidity → belt kinked or taken a set → wrong size |
1. Diagnose in this order, not the other way round
There is a reason the machine comes before the belt. A belt is a consumable that costs a few dollars; the machine settings are free to change and affect every belt you will ever fit. If you swap belts first you have spent money and learned nothing, because the new belt is about to be run in exactly the same conditions that ruined the last one.
The order that resolves problems fastest:
- Machine — contact wheel hardness and face, tension, belt speed, wheel alignment
- Belt selection — is this grain, grit and backing right for this material and this duty?
- Technique — pressure, dwell time, direction of travel, grit sequence
- The belt itself — genuine defects exist, but they are the last hypothesis, not the first
Machine settings are covered in detail in belt grinder setup — speed, contact wheels and tension. What follows assumes you have that page open in the other tab.
2. The belt stops cutting — glazing
What you see: the belt looks fine, the grain is still there, but it slides across the work and generates heat without removing much. Sometimes the surface of the belt has gone visibly shiny.
What is happening: the grain has rounded over instead of fracturing. Zirconia alumina and ceramic grains are designed to break down under load, continuously exposing fresh cutting edges — that self-sharpening behaviour is the entire reason they outlast aluminium oxide on hard metals. Deny them the pressure that causes the fracture and they stop renewing themselves. They dull, and then they rub, and rubbing generates heat, which accelerates the problem.
Where the pressure went:
- Contact wheel too soft — the single most common cause. A soft wheel deflects and spreads the load over a large contact area, dropping unit pressure below what the grain needs.
- Tension too low — a slack belt absorbs the pressure instead of transmitting it into the cut.
- Belt speed too low — each grain gets fewer, gentler engagements per second.
- Operator backing off — often because the belt "felt like it was struggling", which makes it worse.
If the setup is right and it still glazes, the grain may be over-matched to the material — a very hard grain on a soft material will burnish rather than cut. That is a selection problem, not a fault.
3. The workpiece burns or turns blue
What you see: straw, blue or grey discolouration on stainless steel; scorch marks on wood.
What is happening: heat is going into the workpiece faster than the part and the swarf can carry it away. Stainless steel, titanium and high-nickel alloys are all poor conductors of heat, so it stays concentrated exactly where you do not want it.
Causes, in the order they usually apply:
- A dull belt being pushed harder. This is the number one cause. A belt past its useful life converts almost all your input into friction. Change it — the belt was going to be scrapped anyway, and the burnt part may not be recoverable.
- Belt speed too high. More engagements per second means more heat per second.
- Too much removal in one pass. Multiple lighter passes with the part moving, rather than one heavy pass, gives the heat somewhere to go.
- Dwelling in one spot. Keep the belt or the part moving continuously.
Belt-side answers when the process cannot be slowed down: ceramic grain cuts sharply at lower pressure and generates less heat for the same removal, which is why WY1599 is the model we point at titanium, high-nickel and other burn-sensitive work. For dry finishing where heat builds on fine grits, JA513 carries an anti-burn treatment.
4. Evenly spaced marks — chatter
What you see: a regular, repeating pattern across the surface, spaced consistently. The giveaway is the regularity — random scratches have a different cause (see section 5).
What is happening: almost always the joint. It is the thickest and stiffest point on the belt and it strikes the work once per revolution. Behind a hard contact wheel or a rigid platen there is nothing to absorb that impact, so it prints into the surface.
Fixes:
- Move to a softer contact wheel — it deflects at the joint instead of transmitting the impact.
- Reduce tension slightly on the finishing pass.
- Use a belt whose joint has been top-skived. The industry marks this NTS, MTS and FTS for no, medium and full skiving. The finer the grit, the more it matters — on a 400# finishing pass a joint step that is invisible on 60# is the dominant surface defect.
If the spacing does not match one belt revolution, look at the machine instead: worn bearings, an out-of-round wheel, or a workpiece that is flexing under pressure will all produce periodic marks of their own.
5. Scratches deeper than the grit should be making
What you see: random deep lines that a 240# or 320# belt could not have produced.
Cause one — a grit step was skipped. Each grit in a sequence has exactly one job: remove the scratch pattern left by the previous grit. Jump from 80# straight to 240# and the 240# belt physically cannot reach the bottom of the 80# scratches — it polishes the peaks and leaves the valleys, so the coarse scratches survive under a shiny finish and appear the moment the light catches them. Step through the sequence and do not skip more than one grade.
Cause two — contamination. A single coarse grain shed from a previous belt, a chip of hardened swarf, or grit picked up from a dirty bench will drag a line the whole length of the pass. Clean the machine and the part between grit changes, and be especially careful going from coarse to fine.
6. The belt tracks off or wanders
What you see: the belt drifts toward one edge of the wheel, or walks off entirely.
This is usually mechanical. Work through it in this order:
- Tension too low. A slack belt has nothing holding it on line. This is the first thing to check and it resolves the majority of cases.
- Wheels not parallel, or tracking adjustment out. Most machines have a tilt adjuster on the idler; if it has drifted, the belt follows.
- Belt length wrong. A belt slightly too long runs loose regardless of how you set the tracking — it never comes up to working tension. This traces back to how the belt was measured and ordered; see how to measure and order the right belt size.
- Moisture damage. A backing that has absorbed humidity unevenly will distort and pull to one side. See section 9.
- Debris on the wheel. Built-up swarf or resin on the contact or idler wheel changes its effective profile.
7. The belt tears at the joint
What you see: the belt opens up at the splice, usually without warning.
Cause one — fitted backwards. A lap joint is bevelled and overlapped, which makes it directional. Run it against the arrow printed inside the belt and the trailing edge of the overlap catches on the work and peels open. This is the most common cause and it is entirely preventable: check the arrow every time you fit a belt. Butt joints, which are joined with tape or film on the back, are bidirectional and can be reversed deliberately.
Cause two — tension too high. Stress concentrates at the joint because it is the stiffest section. Every MOOSEFOS belt is built with a solid joint specifically to resist breakage, but any joint has a limit and a tensioner wound past what the machine was designed for will find it. If you are breaking belts at the splice, back the tension off before changing supplier.
Cause three — the joint is hitting something. A sharp workpiece edge, a badly set work rest, or a platen with a lip will catch the joint once per revolution until it fails. If the belt lasted a predictable number of minutes each time, suspect an obstruction rather than the belt.
8. The belt clogs and loads up
What you see: material packs into the spaces between grains; the belt looks smeared rather than shiny, and cutting falls off sharply.
What is happening: soft or gummy materials — aluminium, brass, resinous timber, paint and filler — fill the gullets between grains so the abrasive can no longer reach the work. This is a different failure from glazing: in glazing the grain is dull, in loading the grain is buried.
Fixes:
- Use an anti-clog belt. A stearate or similar coating stops material adhering in the first place. JA537 is built for this and runs 180#, 320# and 400#.
- Go coarser. A coarser grit has larger gullets and simply holds more before it fills.
- Raise the belt speed a little so material is thrown clear rather than pressed in.
- Reverse a butt-jointed belt part-way through its life to clear packed material and extend service life. Never do this with a lap joint.
- Match the belt to the metal. For aluminium specifically, see choosing belts for aluminium.
9. The belt was damaged before it ever ran
This is the failure mode most workshops never consider, and it is the one that most often explains "the whole box was bad."
A coated abrasive belt is cloth or paper, resin and grain — all of which respond to their environment. The backing absorbs and releases moisture, and it changes dimension slightly when it does.
- Too humid: the backing swells, the belt loses its flatness and tracks badly, and in bad cases the adhesive bond softens.
- Too dry: the backing becomes brittle and cracks, often at the joint.
- Stored folded, kinked, or leaned against a wall: the belt takes a permanent set and will run with a slap or a wobble that no tracking adjustment fixes.
Practical storage: hang belts on a rack of adequate diameter, or lay them flat. Do not hang them on a nail or a narrow peg, which creates a crease at one point. Keep them in a stable, moderate environment — industry guidance commonly cites roughly 15–25 °C and 45–65 % relative humidity — and away from radiators, exterior doors and direct sun.
10. When it really is the belt
Genuine belt defects do exist, and they have a signature: they are consistent across the batch and independent of the machine. If one belt out of twenty behaves differently, look at that belt. If every belt in the box does the same thing on two different machines, that is worth reporting — and it is worth reporting with detail, because "the belts are no good" cannot be acted on.
What actually helps us diagnose it:
- Model, grit and belt size
- Material being ground and the operation
- Machine speed and driven wheel diameter, plus contact wheel type if you know it
- What the previous belts were, if these replaced something that worked
- A photo of the belt surface and of the workpiece — these two images usually settle glazing versus loading versus burn on their own
11. Common questions
Why has my sanding belt stopped cutting?
If the grain is intact but the belt slides, it has glazed — the grain rounded over instead of fracturing. Check contact wheel durometer, tension and belt speed, in that order, before replacing the belt. Zirconia is affected most because it self-sharpens by fracturing and needs pressure to do it.
What causes chatter marks?
Usually the joint striking the work once per revolution, most visible behind a hard contact wheel or a rigid platen. Move to a softer wheel, ease the tension slightly, or use a top-skived fine-grit belt. If the spacing does not match one belt revolution, suspect the machine instead.
Why does my belt keep tracking off?
Tension too low is the most common reason, followed by wheels out of parallel and a belt that is slightly too long to ever reach working tension. Moisture-distorted backing and swarf built up on a wheel are the less obvious ones.
How should sanding belts be stored?
Hung on a rack or laid flat, never folded or creased, in a stable moderate environment — commonly cited as roughly 15–25 °C and 45–65 % relative humidity. Belts that have shipped through very different conditions should be unpacked and left in the workshop for about a day before use.