How to Read Weld Seam Grinding: the Right Grit

The short answer

The rule that governs nearly every weld seam grinding decision: pick the grit for how much you must remove and how close you are to the finish line, then step the grit down one band per pass so every pass leaves a finer scratch than the one before it. For a typical fillet weld cap on mild steel, that is P60 to P80 to take the cap down, P120 to even out the ground zone, P240 to blend the transition, and P320 to P400 before coating. Match the backing to the demand on the disc, not to the grit: a steel-backing disc for the heavy cap reduction, a fiberglass-backing disc on a fast tool where heat and flex are the problem, and a plastic-backing disc for the light finishing pass. Two scenarios where the standard sequence bends: thin sheet, where a coarse grit on a fast tool builds enough heat to warp the part, so you drop the grit and keep the tool moving; and softer alloys, where P60 cuts deeper than you want, so you start at P120 to P240 instead.

If your job is everyday weld seam grinding on structural steel, the P80 to P120 band on a disc that fits your grinder does most of the real work, and the abrasives glossary keeps the label terms straight while you read the rest of this guide. The full set of diameters and backings we stock is at the flap disc range.

Two habits turn this from guesswork into a repeatable spec: always name the grit standard you are buying to, FEPA or ISO 848, because the same number on two different scales is not the same grain size; and always confirm the arbor hole and the backing against your specific tool, not against what you used before. Everything below is the reasoning behind those two moves.

How a flap disc removes a weld cap

🛠️ Why the disc, and not the wheel. Most of the first and second pass in weld seam grinding happens on a flap disc, and the reason is the construction. A flap disc is a backing plate with dozens of short, overlapping flaps of abrasive cloth arranged in a spiral; as the disc spins, each flap stands up at an angle and the cutting edges keep exposing fresh grit as the old ones wear. That is why a flap disc runs longer and cooler than a single-layer paper disc on the same cap: a paper disc glazes and clogs at exactly the moment the seam asks the most of it, and the flaps roll back and keep cutting instead. In practice, the first time we ran a thick cap on a paper disc, the change in feel at the point it glazed was the lesson that stuck: the cut dies, but the disc looks fine.

Grain decides how the cut behaves; grit decides how much it removes. On the heavy first pass of weld seam grinding, zirconia alumina is the grain to reach for on tough stock, because its tougher, semi open structure keeps the face open when the material is packing the grain. On the blending and finishing passes, aluminium oxide is the forgiving, economical option that gives an even scratch on steel and stainless. Grain decides the feel; grit decides the size of the cut. Change one deliberately and keep the other fixed, or you are solving two problems at once.

The backing is where most of the choice actually happens:

Steel backing ⚙️ — for demanding deburring and cap reduction; it stays rigid under pressure and holds the flaps in place while material comes off.
Fiberglass backing 🔥 — for high-speed work; it resists flex and heat, so it suits fast hand-held grinders where a plastic disc would run hot.
Plastic backing ✅ — for lighter, cooler finishing passes where you want the least pressure and the least heat at the part.

On a thick-plate cap, that is a steel-backing job: you want the rigidity doing the work. On a fast hand-held pass over a fillet, it is a fiberglass job: the disc wants to spin without flexing or overheating. Compare the fiberglass-backing flap disc against the steel-backing flap disc and you will see how the two backings differ in colour and stiffness, which is a useful shop-floor cue before you commit to a grit. One more construction signal: a well-made disc wears in stages, flap by flap, so the cut stays even to the end. If a disc changes character in the middle of a pass, that is a construction quality problem, not a grit problem, and it is worth reporting back to the supplier rather than working around it. Model numbers and stocked backings are on the flap disc range, and the terms for each of the flaps, the backing and the grain are defined in the abrasives glossary.

Picking the grit by material and finish line

The grit number is a FEPA or ISO 848 designation, and the scale is the one that keeps two suppliers from quietly shipping you two different discs: a P120 on one scale is not automatically the same grain size as a 120 on another, so the standard is part of the spec, not a footnote. The reference is at the FEPA grit standard (ISO 848).

Here is the working ladder to keep in your head:

P24 – P40 🛠️ — strip and knock down high material; coarse cap reduction on very proud welds.
P60 – P80 ⚙️ — fast cap reduction; the heavy first pass on most weld seam grinding work.
P100 – P120 🔩 — the everyday grinding and blending band; where most of the real work on mild and stainless steel happens.
P150 – P240 📏 — blend the ground zone into the parent material and prepare the surface.
P320 – P400 ✅ — cleanup and de-raise, leaving a uniform low scratch before coating or inspection.
P1500 and finer ✨ — polish and de-raise on a surface that is already close to its final state.

A worked example: mild-steel fillet, painted finish

P80 takes the cap down to roughly flush. This is where the steel backing earns its place; you want pressure and material movement.
P120 evens out the ground zone and starts blending it into the parent material.
P240 removes the visible line where the weld stops and the plate starts, and prepares the surface for primer.
P400 leaves a uniform, low scratch so the paint film sits evenly and the ground zone does not show through.

The single most common reason a weld looks ground but not finished is a skipped step: P60 scratches sitting under the finish because someone jumped from P60 to P400. The rule is one band at a time. If you are working a P120 abrasive belt on the same kind of seam, the grit logic carries across directly, and the abrasive belt range is where those sizes are listed.

Where the rule bends

Thin sheet 🔥 — a coarse grit on a fast tool builds heat that warps the part. Drop a grit and keep the tool moving; the fix is speed and motion, not grit.
Softer alloys 🟦 — start at P120 to P240 instead of P60, because a coarse grit cuts deeper than the alloy wants.
Coatings already on the part 🧱 — stripping plus grinding in one pass loads a fine grit fast; step the grits and let each pass do its own job.

Material decides the starting grit; the finish line decides the number of steps. The metal fabrication abrasive use notes walk through the sequence by material, and that is the reference to keep open when the job is anything other than a straightforward fillet on mild steel.

Choosing grits by joint type

The same P80-to-P400 ladder does not read the same on every joint, because the geometry changes how much of the disc face actually touches the seam.

Fillet welds 🔩 — the classic case: a proud cap sitting on the intersection of two plates. Grind to flush with the surrounding surface, and the sequence from the worked example above is the one to follow. The cap is the material you are removing, and "flush" is the finish line.
Butt welds on long, straight runs 📏 — on structural frames and beam work, the seam is long and nearly flat, which is where a straight abrasive belt takes the cap in one pass along the length, and the disc does the transition blending and cleanup. That belt-then-disc combination is the pattern a shop running P120 abrasive belts uses on frame and beam work, and the metal fabrication abrasive use page shows it in context.
Pipe and curved seams ⭕ — the disc works on a curve that wants to dig deeper than a flat pass does. Drop a grit so the cut is gentler, keep the tool moving around the curve, and let the smaller-diameter disc reach the seam. A 100 mm or 115 mm disc gets into places a 180 mm disc physically cannot, and the guard has to suit the diameter you actually fit, which is the point to confirm against the tool before the first pass.
Stainless of any joint 🧊 — the material rule stacks on top of the geometry rule: keep the disc dedicated to stainless, or change grits and clean the tool between a steel job and a stainless one. Embedded mild-steel particles in the grain are the seed of rust spots on the stainless part later, and the stainless steel finishing fundamentals page is a plain reference for why the corrosion story works that way. Where the seam has to de-raise before a clear coat, a 1200 grit is the top pass. For the seam itself, the stainless steel finishing use notes give the grit sequence adapted to the metal.

When two joints share a bench, the dedicated-disc rule is the one that saves the most rework, because a rust spot discovered after coating is a far more expensive fix than a grit swap before the pass starts.

Reading the seam before you commit to a grit

Before the disc touches anything, read the weld the way a fabricator would, because what you see on the seam tells you where to start the grit ladder and how many bands the job will take.

Proud, cap-heavy seam 🛠️ — the cap sits well above the plate. That is a two-pass job: a coarse band to take the cap down, then a middle band to even it off. If the cap is very high, a P40 to P60 opener is justified, but only for that first strike.
Nearly flush seam ✅ — the welder left it close. You do not need the coarse band at all; P120 straight into P240 and P400 is enough, and starting coarse here would dig past the cap and into the parent material.
Undercut at the toe ⚠️ — a groove cut into the plate where the weld meets the parent metal. This is not excess material to remove but a stress riser to fill. Grinding the toe rounds it off, but a deep or sharp undercut often means the weld needs repairing, not just blending, so stop and raise it before you grind.
Porosity and spatter 🧱 — pitted surface and spatter beads riding on top. Spatter is the easy part, a coarse pass knocks it off; porosity is harder, because the pits hold the disc back and the fine grit cannot reach into them, so you may need to open the face coarse first and then work down.
Heat-affected colour 🧊 — on stainless, blue or grey discoloration along the seam is an oxidation layer, not a scratch pattern. The grit removes the oxide, but the finish you are aiming for decides how fine you have to go, because a light oxide film needs no more than a fine band, while heavy, dark oxide wants a coarser start.

The single habit this gives you is to count the bands before you buy the grits. A proud cap is four bands of work; a flush seam is two. Reading the seam first is what stops a shop from grabbing one "universal" grit and spending an hour blending, or from burning through a fine disc on a job that never needed it. The metal fabrication abrasive use notes match these seam types to the grit sequence when the job is anything but the ordinary fillet, and the term definitions are in the abrasives glossary.

When to reach for belts instead of discs

A straight seam on a flat or nearly flat surface is where a belt beats a disc, and the deciding factor is the shape of the work, not the grit: disc grits and belt grits run on the same FEPA scale, so the P120 you specify for the disc is the P120 you can spec on the belt, which is what makes a belt-then-disc sequence one spec sheet instead of two.

533 × 75 mm — the size that works on hand-guided sanders over a medium-length seam; light, fast, and easy to keep moving.
610 × 100 mm — the workhorse for bigger frames and longer straight runs on a hand sander.
1520 × 630 mm and 2200 × 1330 mm — dedicated-sander sizes for the long, straight weld runs where the belt is doing the heavy removal along the whole length.

All four are aluminium oxide belts, and they are where the heavy first pass on a straight seam should happen if the seam is long enough to earn it. For curved or tight-radius seams, the flaps of a disc reach the geometry that a belt face cannot, so the disc keeps its job. Browse the full set, including grits from P60 to P5000, at the abrasive belt range, or the all abrasive products page when the job needs both a belt and a disc in the same spec.

Grain and backing, matched to the metal you are welding

The grit ladder sets how much comes off in each pass; the grain and the backing set how the pass feels, and the two choices should follow the metal and the speed of the tool, not the price on the shelf.

Aluminium oxide is the default on carbon and mild steel, because it cuts predictably and loads slowly, so the face stays open for a long seam. It is the grain to reach for when the seam is clean and the cap is the main thing to remove.
Zirconia alumina is the pick when the seam is dirty or the cap is hard, because its tougher, semi open structure shrugs off the packing that shuts a softer grain down. It runs hotter, which pairs with a steel or fiberglass backing on the heavy pass.
Silicon carbide is what you reach for on non ferrous and softer alloy work, where a harder grain keeps the scratch clean instead of glazing. The property behind that is at the properties of silicon carbide, and the everyday aluminium oxide default is at the properties of aluminium oxide.
Ceramic is the high-volume option for a shop that grinds the same seam type all day, because the cost per finished part is what gets managed there, not the price per disc.

Match the backing to the speed of the tool the same way. A fast hand-held machine that runs the seam at high RPM wants a fiberglass backing, because it resists the flex and heat that make a plastic disc wash out on a fast pass. A steady machine at moderate speed on a heavy cap wants steel, because the rigidity is doing the work. The two stocked options sit side by side on the fiberglass-backing flap disc and the steel-backing flap disc pages, and that is where you can check the diameters against the tool before you commit to a grit.

The rule that keeps the three of them, grain, grit and backing, from fighting each other: match the grain to the material, then set the grit to the cut, and only then pick the backing for the speed. Change one of the three at a time and note what moves, because most of the "the disc is wrong" complaints on the floor are really a grain that is out of line with the metal, not a grit problem at all.

A worked enquiry: what to send the supplier

The difference between an order that arrives right and one that needs chasing is the fields you put in the enquiry, not the price you negotiate. Here is what a complete one looks like for a steady grinding job:

Grain and grit: zirconia alumina, P80 and P120, FEPA (ISO 848) scale — because the P80 does the cap reduction and the P120 does the blend, and the scale is named so the two suppliers you buy from ship the same grain size.
Backing: steel backing on the P80 for the heavy pass, fiberglass on the P120 if the tool runs fast — the backing follows the demand on the disc, not the grit.
Diameter and arbor: 125 mm with the arbor your grinder is built for, confirmed against the flange you actually run, because the arbor is the number people forget and the one that has to be right.
Maximum working speed: the model number of the disc you are quoting, with the speed confirmed against it, because the rating comes from the disc, not from a habit.
Quantity and what it ships as: the run quantity and whether it needs to leave as stock or as a branded run with your label, which sets the lead time.

Send those five fields and the reply is a conversation instead of a gamble: the supplier can pre-select the grit and the backing, and you can compare the cost per finished seam rather than the sticker price. The OEM & private-label service page walks through the branded-run fields in more detail, the all abrasive products index shows what is in the range, and the request a quotation form is where the five fields get typed in. If you are new to ordering against a spec sheet, ask for a sample against the spec before the full order, so the grit, the backing and the label all match what the shop expects on the bench.

Speed, dust and safety, in plain terms

⚙️ The one number that protects the operator. Every disc carries a maximum working speed, and the tool must stay at or below it. Surface speed rises with both diameter and RPM, so the same grinder that is fine with a 100 mm disc may be too fast for a 180 mm disc on the same machine. Never assume the tool is within the disc's rating; run it at no load, check the speed, and keep the tool under the disc's printed maximum. No-load speeds on hand tools sit in the thousands of RPM, so the disc's printed rating is the number to check, not your habit. If the number is not on your label, confirm the rated speed with sales and keep the tool below it, because that number has to come from the disc, not from a guess.

The guard is the right size for the disc and it is on, every pass.
The arbor and flange match the disc hole, or a rated adapter is doing the matching, never a modified hole.
Eye and face protection at the point of work, and respiratory protection matched to the dust load.
Dust extraction actually running at the point of work, not just sitting nearby.

🧊 Why dust deserves its own line on a seam. A proud weld cap is hard, cold-worked metal, and it throws more fines than the surrounding parent material, so weld seam grinding generates more dust than the job looks like it should. On coated work you are also removing coating residue along with the metal, and the dust control has to plan for that load. The U.S. PPE general requirements are a useful baseline at OSHA 29 CFR 1910.132, and a practical reading of how to set up and use an abrasive wheel safely is at CCOHS: Abrasive Wheels. Read both before the team starts, and re-check the guard and the arbor each shift.

Two failure modes are worth naming specifically. A disc run past its speed rating can separate at the flaps, and a disc seated on a mismatched flange can walk and catch; both are preventable, and both start at the label. The habit that covers them is short: read the disc, read the tool, and only then start the job.

Cost, life and lead time

On steady volume, the number that matters is cost per finished seam, not cost per disc. A disc that runs longer and cooler cuts more before it has to come off, so the per-seam cost on a good grade can land lower than on a cheap one, and the difference shows up in how many seams you get between disc changes. A concrete way to think about it: if one disc costs 30 percent more but lasts three times as long over the same seam, the per-seam cost is far lower on the pricier disc. That is the number to ask a supplier to help you with, and it is the number a bulk conversation should be built around.

The fields that make or break a bulk or branded order are the same ones on every spec sheet, and they are exactly what the OEM & private-label service page walks through:

Grain and grit — the grain, and the FEPA or ISO 848 grit, named on the same scale you shop to.
Backing — steel, fiberglass or plastic, chosen for the speed and the job.
Diameter and arbor — matched to the tool and flange you actually run.
Maximum working speed — confirmed against the model, never assumed.
Label and packing — your logo, your part code, any spec sheet you want on the box for a private-label run.
Quantity and lead time — which sets whether the order ships from stock or gets made to your run.

On lead time, two things set it: in-stock grits and common diameters usually move fast, and off-grid grits or a backing that does not run in the same batch take longer, because the flaps, the backing and the grain all have to line up. A branded run adds time for the label and the packing, and the minimum order quantities and exact lead times are confirmed against your spec during the request a quotation conversation.

On wear, the pattern to expect is: a steel-backing disc on an aggressive seam runs down the flaps first, which is normal, because the flaps are the wearing part and the backing is what holds them. If a plastic-backing disc on a fast finishing pass is running hot, the life is limited by the temperature of the job, not by how hard you press, and the fix is the backing, not the grit. If your life is short and the disc feels hot at the end of a pass, you are on the wrong backing for the speed, not the wrong grit.

When a flap disc stops cutting

The complaint is usually that the disc looks fine and the machine sounds fine, but the seam is not going. The causes split cleanly, and each has its own fix:

Loading. The grain face is packed with soft material. Step to a grit that keeps the face open, change to a tougher grain if the material is the cause, and keep the disc moving.
Glazing. The grit face has flattened and stopped biting. Step up a grit or change the grain, and the cut comes back.
Grit too fine for the job. A finishing grit on a heavy seam asks the disc to polish instead of remove. Step down a grit and the cut returns.
Tool speed too low. Below the speed the disc was designed for, it cuts soft and slow. Check the no-load speed and keep the tool in the range the disc was made for.
Pressure instead of grit. If the cut dropped when you leaned in, the fix is lighter pressure and a moving disc, not a coarser grit.

Under cutting is almost always a match problem, not a bad disc: the grit, the grain, the backing or the speed is out of line with the job, and fixing that one match is what brings the cut back. The abrasives glossary keeps the terms for each of those four aligned, so you and your team are diagnosing the same problem rather than each blaming a different one.

Common mistakes and how to fix them

The fixes below are cheaper than the mistakes they prevent, and each one shows up on the floor more often than you would expect.

Skipping grits 📏 — going from P60 straight to P400 leaves the P60 scratches under the finish. Add the P120 and P240 steps and the surface clears.
Wrong arbor 🔩 — a disc that does not sit fully on the flange walks at speed. Use the right hole or a rated adapter, and never file the hole larger.
Running the tool too fast ⚙️ — if the disc is below the tool speed, lower the tool speed or use a smaller diameter, and never run a disc past its printed maximum.
Loading and glazing ✅ — the grain face packs with soft material or the grit face flattens. Step the grit, change the grain, and keep the disc moving instead of forcing it into the work.
Heat on a thin part 🔥 — a coarse grit on a fast tool warps thin sheet. Drop a grit and keep the tool moving rather than pressing in.
Shared discs across metals 🧊 — using the same disc on a steel job and then a stainless one embeds mild-steel particles in the grain, and those are the seed of rust spots on the stainless later. Keep the disc dedicated, or swap grits and clean the tool between the two metals.

The single thread through all six is that the disc is only as good as the grit, the backing, the arbor and the speed you matched to the job before it started; most of the complaints on the floor are mismatch complaints wearing a grit's name. The metal fabrication abrasive use notes and the abrasive belt range are where to confirm the right grit and backing for the specific job before the next batch runs.

FAQ

What grit do I start on for general weld seam grinding? Start at P80 to P120 on steel for the cap reduction and blending, and step down to P240 or P400 for the finish before coating. One band at a time, and never skip two grits in a row.

I skipped a grit and now the surface shows coarse scratches. What gives? The coarser scratches sit under the finish because the skipped band never ran. Work back through the grit you skipped, one band at a time, and the surface will clear. The fix is the sequence, not a new disc.

When is a belt better than a disc on the seam? On a straight seam on a flat or nearly flat surface, a belt takes the cap in one pass along the length, and the disc does the transition blending and cleanup. On curved or tight-radius seams, the disc reaches the geometry a belt face cannot, so the disc keeps its job there. That combination is what frame and beam work settles into in practice.

Why did my flap discs stop cutting on the last few seams? It is one of four things: loading, glazing, a grit that is too fine for the job, or a tool speed that is too low. Each has its own fix, and none of them is "buy a better disc." Check the grain face, the grit and the tool speed in that order.

What changes when the seam is on stainless steel? Two things: keep the disc dedicated to stainless so mild-steel particles never reach the grain, and use the finer finishing grits so the scratch pattern the finish requires is in place. The stainless steel finishing use notes carry the full sequence for the metal.

How do I buy flap discs in bulk for a regular job? Send the grit, the backing, the diameter, the arbor and the grit standard to a supplier, and ask for cost per finished part, not cost per disc. The flap disc range is where the grit and backing options are listed, and the request a quotation form is where a bulk or branded order starts.

What standard should I name in an enquiry? Name FEPA or ISO 848, plus the diameter and the arbor, so your order is not mixed with a different grit scale. The reference is at the FEPA grit standard (ISO 848), and that is the scale to hold your shop to across every supplier. If the finish line is a very light de-raise rather than a coating, a 1200 grit on a plastic or fiberglass-backing disc is the band to ask for in that enquiry.

Before you buy

A short compliance and purchasing checklist to run before the first pass and before the next order. This is a reminder, not a substitute for your own site rules, and under compliant use with the right protection the risk of these tools is reduced but not zero.

Confirm your tool's maximum speed is at or below the disc's rated maximum working speed before every use.
Check the arbor hole and the flange are the correct size, and use a rated adapter where the two need matching.
Fit the guard, the right size for the disc, and keep it on for every pass.
Wear eye and face protection, and use dust extraction and respiratory protection matched to the dust load at the point of work.
Name the grit standard, FEPA or ISO 848, in every order so the scale is never the quiet mismatch.
Check the certificate of conformity and any inspection report against your spec before you accept a bulk lot, especially on a branded run.
Store the stock dry, cool and away from moisture, and confirm transport and packing conditions for a long haul.
Follow the maker's install notes, and check the part and the disc for heat and scratch depth on thin or soft material.
Confirm lead time, minimum order and packing for a bulk or private-label run, and start that conversation on the request a quotation page.

For the standards and setup references behind most of the items above, the U.S. baseline is OSHA 29 CFR 1910.132 — PPE general requirements and the practical setup guide is CCOHS: Abrasive Wheels. The manufacturing and quality controls behind the stock you are buying are on the manufacturing & quality page.

If you tell us the material, the finish line you are aiming for and the tool you run, we can pre-select the grit and the backing to start from, and work the full sequence with you on the request a quotation page.