Process & DFM
Deburring Methods Compared
The short answer
Every cutting operation leaves a burr, so deburring is not optional, only unspecified. Hand deburring is universal and scales badly. Tumbling and vibratory finishing are cheap per part at volume but round every edge indiscriminately. Thermal and electrochemical methods reach internal passages nothing else can. The method should follow from your part geometry and quantity, and your drawing should state an edge condition rather than leaving it to be guessed.
Why burrs exist and why they cost you
A burr is displaced material, not removed material. As a cutting edge leaves the workpiece it pushes rather than shears, and the metal folds over the edge instead of coming off as a chip. Ductile materials burr worse than brittle ones, which is why aluminum and low carbon steel burr more than cast iron.
That means burrs are a consequence of cutting, not a defect. Every hole, every slot, every profile leaves them, and somebody has to deal with them before the part is safe to handle and fit to assemble.
The cost is real and usually invisible on the quote. On a part with many holes and intersecting features, hand deburring can rival the machining time. It is also almost entirely labor, which means it does not fall with volume the way machining time does unless you change the method.
The expensive version is the part that arrives with sharp edges because the drawing said nothing, and gets deburred by your own team on receipt.
The methods, compared
Match the method to geometry and quantity rather than to habit. The finish column matters as much as cost, because several of these change the part beyond removing the burr.
| Method | How it works | Best for | Watch out for |
|---|---|---|---|
| Hand deburring | Files, scrapers, abrasive pads | Prototypes, low volume, selective edges | Labor cost per part is flat, operator to operator variation |
| Machine chamfer or back spot face | A tool run in the same setup | Any edge the cutter can reach | Adds cycle time, cannot reach cross holes |
| Vibratory or tumble finishing | Parts and abrasive media in motion | High volume small parts | Rounds every edge, can round features you needed sharp |
| Bead or abrasive blasting | Media propelled at the surface | Cosmetic uniformity, light burrs | Does not remove substantial burrs, changes the surface |
| Brush deburring | Abrasive filament brushes | Flat faces and top edges at volume | Poor access into holes and pockets |
| Thermal energy method | A gas pulse burns off burrs | Complex internal passages, hydraulic parts | High setup cost, only economic at volume |
| Electrochemical deburring | Anodic dissolution at the burr | Cross holes, internal intersections | Tooling per feature, requires conductive material |
| Abrasive flow machining | Abrasive putty forced through passages | Internal channels, manifolds | Expensive, alters passage dimensions |
The edge you actually want, and how to say it
Most drawings say nothing about edges, or write a blanket note like break all sharp edges. That note is better than silence but still leaves the size open, and edge condition is a dimension like any other.
A useful default is a stated range: break edges 0.1 to 0.3 mm, or 0.005 to 0.015 inch. That tells the shop the intent and gives them room to hit it with an ordinary tool.
Where an edge matters, call it out specifically. A chamfer that guides an O ring past a bore, a lead-in on a press fit, an edge a hand will touch, and a corner a coating has to cover uniformly all deserve their own callout with a size and tolerance.
Equally important is saying which edges must stay sharp. Sealing lands, mating faces, knife edges and datum surfaces can be ruined by a tumble that rounded everything. If you send a part to bulk finishing without protecting those, you will get them back rounded.
| Edge callout | Means | Use when |
|---|---|---|
| Break sharp edges | Remove the burr, size unstated | Non critical edges, prototypes |
| Break edges 0.1 to 0.3 mm | A defined, inspectable range | Default for production drawings |
| Chamfer 0.5 x 45 degrees | A dimensioned feature, machined | Lead-ins, seal entries, fastener seats |
| Radius R0.5 max | A rounded edge with an upper bound | Coated parts, handled surfaces |
| Sharp edge permissible | Explicitly do not break | Sealing lands, knife edges, datums |
| No burr, edge condition per ISO 13715 | Formal edge state specification | Regulated or safety critical parts |
How volume changes the answer
At one to ten parts, hand deburring is correct and any discussion of tumbling is theoretical. The setup for bulk finishing exceeds the labor of doing it by hand.
Between roughly fifty and a few hundred parts, the calculation shifts. Vibratory finishing amortizes its setup, and the per part labor drops sharply. This is usually where a shop will suggest it, and where you need to have already told them which edges must stay sharp.
At production volume the real answer is often to design the burr out rather than to remove it faster. A back spot face on a cross hole, a chamfer added in the same setup, or moving a hole so it does not break into a pocket wall can each eliminate a manual operation permanently.
That is the highest leverage move available, and it is a design change rather than a process change, which is why it belongs in the DFM conversation rather than the quoting one.
The burrs nobody sees until assembly
Cross hole intersections are the classic failure. Where one drilled hole breaks into another, the burr forms inside the part, where no hand tool reaches and no visual inspection finds it. On a hydraulic or pneumatic part that burr breaks loose later and travels into something expensive.
Deep pockets with tight corners are the second. The burr forms at the floor to wall intersection and a deburring tool cannot get at it without damaging the wall.
Threaded holes are the third: the burr at the top of a tapped hole interferes with the fastener seating flat, which quietly changes the preload on an assembly nobody thought was tolerance sensitive.
If your part has any of these, say so when you send it, and expect the answer to involve either a dedicated process such as thermal or electrochemical deburring, or a design change that removes the intersection.
What this costs you
Deburring is labor, and labor is the line item that does not shrink with volume unless the method changes.
On a typical machined part with a moderate number of features, hand deburring commonly runs a meaningful fraction of the machining time, and on hole-rich parts it can approach or exceed it. Because it is quoted as part of the operation rather than separately, most buyers never see how much they are paying for it.
The two ways to reduce it are both cheap to apply. First, specify the edge condition so the shop is not guessing at a standard higher than you need; a mirror-uniform hand finish costs multiples of a functional break. Second, remove the burr sources in the design.
The way to increase it accidentally is a blanket note demanding all edges deburred and polished on a part with fifty holes. That is a specification of labor, whether or not it was meant as one.
What to send, and what to ask
Send the model with an edge note that states a range rather than an intention, and mark the edges that must stay sharp. If any hole breaks into another hole or into a pocket wall, flag it explicitly.
Give your quantity now and your expected annual quantity, because the right method at ten parts is the wrong one at a thousand.
Then ask three questions. How are you deburring this part, and is that included in the price? Which features here are driving the deburring time? And is there a chamfer or a hole position I could change that would remove a manual operation entirely?
That third question is the one that pays repeatedly, because the answer applies to every unit you ever order.