The process that’s right for one prototype is often the wrong process for 5,000 units — and the reverse is just as true. Knowing when to switch is a cost decision, not a maturity milestone.

Quick summary: CNC machining and 3D printing dominate prototyping because they require no tooling investment. Injection molding and other tooled processes dominate production because per-unit cost drops dramatically once tooling is amortized — but only past a volume threshold that justifies the tooling cost.

Why Is CNC or 3D Printing Usually Right for Prototypes?

No tooling investment. You can iterate a design multiple times without paying for a mold each time you change a feature.

Why Does Injection Molding Take Over at Higher Volumes?

Mold tooling is expensive upfront, but the per-unit cost afterward is very low. Past a certain volume, that trade pays for itself — often faster than most people expect.

How Do You Know When You’ve Crossed the Threshold?

Compare total cost at your expected volume: CNC/3D print cost scales roughly linearly per unit; injection molding has a large fixed cost plus a low variable cost. The crossover point depends on part complexity and material, but it’s a calculable number, not a guess.

What Happens If You Switch Processes Too Early or Too Late?

Switching to tooling too early risks paying for a mold before the design is finalized — expensive if it changes. Staying on CNC too long into production means paying a higher per-unit cost indefinitely.

FAQ

Can a design go straight from sketch to injection molding, skipping CNC prototyping? Technically yes, but it’s risky — any design flaw discovered after the mold is cut is expensive to fix.

Is there a hybrid approach? Yes — some programs use CNC-machined “bridge tooling” to produce a few hundred units while production tooling is still being built.

Related: Injection Molding Services

Every individual part can measure perfectly within its own tolerance and the assembly still won’t go together — because tolerance stack-up isn’t about whether any one part is correct, it’s about what happens when several correct parts’ tolerances compound in the same direction.

Quick Summary

• Tolerance stack-up is the cumulative variation that builds up across a chain of mating dimensions in an assembly, not the tolerance on any single part.
• Every part in the stack can be perfectly in-spec on its own and the assembly can still fail to fit, if the tolerances happen to compound in the same direction.
• Worst-case stack-up analysis (adding every tolerance at its limit) is the most conservative approach; statistical methods (like root-sum-square) reflect real-world variation more realistically but require more analysis.
• The more parts in a mating chain, the more the stack-up matters — a two-part assembly is a much smaller risk than a five-part stack.
• GD&T with proper datums, rather than a chain of independent linear tolerances, is one of the most effective ways to control stack-up directly.

What Is Tolerance Stack-Up, Actually?

When several parts mate together — a shaft through a series of spacers into a housing, for instance — each part’s individual tolerance contributes to the total variation in the final assembled dimension. If every part in that chain happens to land at the high end of its tolerance range, the total variation can be significantly larger than any single part’s tolerance alone, even though every part individually passed inspection. eMachineShop’s own mating parts guidance puts the core rule plainly: the lowest possible dimension of the female part needs to stay greater than the largest possible dimension of the male part, accounting for both parts’ tolerance ranges at once, not just their nominal sizes.

Why Can an Assembly Fail Even Though Every Part Is In-Spec?

Because “in-spec” only guarantees that one dimension falls inside its own tolerance band — it says nothing about how that variation interacts with the variation in the next part, and the part after that. A five-part stack where each part is off by a few thousandths in the same direction can add up to a gap or interference that no individual inspection would catch, because no individual part actually failed its own inspection.

What’s the Difference Between Worst-Case and Statistical Stack-Up Analysis?

Worst-case analysis assumes every part in the chain lands simultaneously at its tolerance limit, in the direction that makes things worst — it’s the most conservative approach and guarantees the assembly works even in an unlikely worst case, but it can force tighter, more expensive tolerances than statistically necessary, since that exact worst-case combination is often statistically rare. This is exactly why it’s worth knowing what the standard default tolerance actually is before assuming every part needs a custom, tighter spec — the baseline is often adequate for most of the parts in a stack, with only the true fit-critical dimensions needing anything tighter.

Statistical methods, like root-sum-square, instead treat each part’s variation as following a distribution and calculate the likely combined variation rather than the absolute worst case. This usually allows for looser, cheaper individual tolerances while still keeping the assembly reliable in the vast majority of real cases — the tradeoff is that a small percentage of assemblies could theoretically still fail, which needs to be an acceptable risk for the application.

How Does the Number of Parts in the Chain Change the Risk?

Directly and significantly. A two-part mating interface has a much smaller stack-up risk than a five-part chain, simply because there are fewer tolerances compounding. This is one of the underappreciated reasons to consolidate parts in an assembly where possible — fewer mating interfaces means fewer opportunities for tolerances to compound into an interference or gap problem.

How Does GD&T Help Control This?

A chain of independent linear tolerances lets each part’s dimension wander independently, with no explicit control over how features relate to each other across the assembly. GD&T, properly applied with clear datums, controls the relationship between features directly — position, concentricity, parallelism relative to a shared reference — which constrains how the variation can combine rather than leaving every part free to drift independently in whatever direction its own tolerance allows. Every dimension having a tolerance, and tolerances applying to the full length of a feature rather than just a sample point, are foundational GD&T rules that exist specifically to close the ambiguity gaps a plain linear-tolerance chain leaves open.

Frequently Asked Questions

Does tolerance stack-up only matter for precision assemblies? No — it applies to any assembly with more than one mating part, though the consequences are more visible in precision work. A loose consumer product might absorb some stack-up without anyone noticing; a bearing assembly or optical mount won’t.

Is a tighter tolerance on every part in the chain always the safest fix? It’s the most conservative fix, but not always the smartest one. Tightening every part’s tolerance drives up cost across the whole chain; a stack-up analysis often reveals that only one or two dimensions in the chain actually need to be tightened to solve the problem, while the rest can stay at standard tolerance.

How do I know if my assembly needs a formal stack-up analysis rather than just tolerancing each part reasonably? If the assembly has three or more mating parts in a critical dimensional chain, or if a failure to fit would be expensive or unsafe, a formal analysis is worth the time. For a simple two-part interface with generous clearance, reasonable individual tolerancing is often sufficient without formal analysis.

Can a stack-up problem be fixed after parts are already made, or does it require a redesign? Sometimes a shim, spacer, or adjustable feature can absorb the accumulated variation without redesigning the mating parts themselves — this is a common practical fix. But it’s cheaper by far to catch the stack-up risk during design than to add a compensating feature after parts are already in production.

Working through a multi-part assembly and want a second opinion on where tolerance stack-up might bite? Upload your drawings and we can talk through the mating dimensions before they’re locked in — our pre-order checklist and cost-reduction guide are both worth a pass first, since “use the loosest applicable tolerance” is exactly the instinct that keeps a multi-part stack-up problem from becoming a multi-part cost problem too.

Written by James Wright, eMachineShop

ITAR controls who is allowed to access certain defense-related technical data and hardware. CMMC governs how that data has to be protected once you have access to it. They’re related, but they answer different questions.

Quick summary: ITAR (International Traffic in Arms Regulations, 22 CFR Parts 120–130) is an export-control regime restricting access to defense articles and technical data to authorized U.S. persons and registered entities. CMMC (Cybersecurity Maturity Model Certification), built on the NIST SP 800-171 security controls, governs how Controlled Unclassified Information must be protected in a contractor’s IT environment. A company can be ITAR-registered without being CMMC-certified, and vice versa, depending on what data it handles.

What Does ITAR Actually Control?

ITAR restricts the export — including disclosure to foreign persons — of defense articles and associated technical data. Registration under ITAR is a prerequisite for legally handling that category of information at all.

What Does CMMC Actually Control?

CMMC is a cybersecurity certification framework, built on NIST SP 800-171, that verifies a contractor’s IT systems and processes meet defined security-control levels for protecting Controlled Unclassified Information (CUI).

Can a Company Have One Without the Other?

Yes. ITAR registration is about legal eligibility to handle defense technical data. CMMC is about whether your systems are secure enough to protect that data once you have it. A company can be ITAR-registered while still working toward CMMC certification, or vice versa depending on contract requirements.

Why Do Both Matter to a Customer Choosing a Machine Shop?

If a program requires a shop to receive ITAR-controlled technical data, the shop needs both — the legal eligibility under ITAR and, increasingly, the security posture required by CMMC — to be a viable long-term partner.

FAQ

Does ITAR registration automatically mean CMMC compliance? No — they’re governed by different regulations and evaluate different things (legal eligibility vs. cybersecurity posture).

Which one applies to unclassified technical data specifically? Both can, depending on the data’s classification — ITAR governs export control of defense articles/data broadly, while CMMC governs protection of CUI specifically.

Sources: 22 CFR Parts 120–130 (ITAR); NIST SP 800-171 (basis for CMMC controls). Related: eMachineShop is ITAR Registered and JCP Certified — see Our Company.

A CNC end mill is round. It cannot cut a perfectly square internal corner — physically impossible with a rotating cutting tool. Every internal corner on a machined part carries a radius equal to at least the tool’s radius.

Quick summary: Internal corners on CNC-machined parts always carry a fillet radius determined by cutter diameter. Designing to accept this — rather than specifying a “sharp” internal corner — avoids costly secondary operations like EDM or hand-finishing.

Why Does the Tool Geometry Force a Radius?

A ½” end mill can’t cut a 0” internal radius — the corner will always show the arc of the tool. Smaller tools produce smaller radii but cost more time and risk breakage on deep pockets.

What Happens If the Drawing Calls for a Sharp Internal Corner?

The shop either flags it back to you, or machines it with a smaller tool (slower, costlier) or a secondary EDM operation (costlier still) to achieve a true sharp corner.

What’s the Practical Fix?

Add a corner relief or design the mating part with a slight chamfer or radius that accommodates the machined fillet. This is standard design-for-manufacturability practice and avoids the cost entirely.

How Do You Specify This Correctly on a Drawing?

Call out a maximum allowable internal radius rather than “sharp corner.” Per ASME Y14.5-2018, an explicit radius callout removes ambiguity about what the shop should target.

FAQ

Is there a minimum achievable internal radius? It depends on pocket depth and available tooling — deeper pockets typically require larger-diameter, more rigid tools, which means a larger minimum radius.

Does this apply to external corners too? No — external corners can be machined genuinely sharp; it’s internal corners that inherit the tool radius.

Source: ASME Y14.5-2018. Related: GD&T Reference Guide

If your part is flat and 2D, laser or waterjet is often faster and cheaper than CNC. If it needs 3D features — pockets, threaded holes, contoured surfaces — CNC is usually the only option of the three.

Quick summary: Laser cutting is fastest and cheapest for thin sheet metal with simple 2D geometry. Waterjet handles thicker material and heat-sensitive metals without a heat-affected zone. CNC machining is required for any part needing 3D features, tight tolerances, or threaded holes.

When Is Laser Cutting the Right Call?

Thin sheet (typically under ½”), simple 2D profiles, high volume. Fast cycle times keep cost down, but laser introduces a heat-affected zone that can matter for some materials.

When Does Waterjet Make More Sense Than Laser?

Thicker material, heat-sensitive alloys (titanium, some stainless grades), or parts where a heat-affected zone isn’t acceptable. Waterjet is slower than laser but avoids thermal distortion entirely.

When Do You Actually Need CNC?

Any 3D feature — a pocket, a counterbore, an internal thread — takes CNC out of “one of three options” and into “the only option.” Laser and waterjet are 2D-cutting processes; CNC is not.

Can These Processes Be Combined?

Often, yes — a part can be laser-cut or waterjet-cut to a rough profile, then CNC-finished for tight-tolerance features. This can be cheaper than machining the whole part from solid stock.

FAQ

Is waterjet always more expensive than laser? Usually, per part — but not always per project, especially on thick or heat-sensitive material where laser would introduce distortion.

Can laser cutting hold tight tolerances? Not to CNC-level tolerances on most materials — laser is better suited to profile cutting than precision fit features.

Related: Laser Cutting Services · Waterjet Cutting Services

A $40 bracket and a $400 bracket can look nearly identical on a drawing. The difference almost never comes down to greed or an inflated shop rate — it comes down to ten specific variables, and most of them are decisions the designer made without realizing they had a price tag attached.

Quick summary: CNC part cost is driven by material, tolerance, surface finish, part geometry, setup complexity, quantity, machine time, secondary operations, inspection requirements, and lead time. Tightening any one of these beyond what the part actually needs adds cost without adding function. Below is what each factor does to your quote, and where the “expensive for no reason” costs usually hide.

What Material Are You Actually Specifying?

Raw material cost varies enormously — 6061 aluminum and titanium can differ by 10x per pound, and machinability differs almost as much. Titanium and hardened steels run slower spindle speeds, wear tooling faster, and need more frequent tool changes, all of which show up as machine time on your quote.

How Tight Are Your Tolerances, and Do You Need Them That Tight?

This is the single biggest avoidable cost driver we see. Under ASME Y14.5-2018, the current U.S. standard for dimensioning and tolerancing, a tolerance is a functional requirement, not a precision flex — it exists to guarantee fit and performance, nothing more. A ±0.005” tolerance is achievable on most standard 3-axis work at normal speed. Push that to ±0.0005” and you’re often looking at secondary grinding, more frequent in-process inspection, and multiple setups just to hold the number — on a feature that may not have needed it. Before specifying anything tighter than ±0.005”, it’s worth asking whether the assembly actually requires it or whether it’s a default carried over from a previous drawing.

What Does Your Surface Finish Callout Actually Cost?

A standard machined finish (usually 125–250 Ra) comes off the tool with no extra steps. Anything smoother — a 32 Ra finish for a sealing surface, for example — usually means a secondary grinding or polishing pass. That’s not a line-item nuisance; it’s real added machine time and, often, a different piece of equipment entirely.

How Complex Is the Part Geometry?

Deep pockets, thin walls, internal features reachable only from odd angles — all of these slow the toolpath down and increase the risk of a scrapped part mid-run. Geometry that requires a 5-axis setup instead of 3-axis, or multiple work-holding changes, adds setup time that gets amortized into the unit price.

How Many Setups Does the Part Require?

Every time a part has to be flipped, re-fixtured, or moved to a different machine, that’s non-cutting time the shop still has to charge for. A part designed so most features can be reached in one or two setups is meaningfully cheaper than one that needs four.

What Quantity Are You Ordering?

Setup and programming costs are fixed regardless of whether you order 1 part or 100. At low quantities, you’re paying almost entirely for setup; at higher quantities, that fixed cost spreads out and the per-unit price drops. This is why a 10-piece order and a 100-piece order of the same part rarely scale linearly — see our companion piece on quantity pricing for the specific breakdown.

How Much Material Are You Removing?

A part machined from a large block down to a small final shape — high material removal — takes longer to cut and wastes more raw stock than a part closer in size to its starting stock. When possible, starting from bar stock sized closer to the finished part reduces both machine time and scrap.

Do You Need Secondary Operations?

Anodizing, powder coating, heat treatment, tapping, and similar operations are often outsourced to specialty vendors, which adds both cost and lead time on top of the base machining price. These aren’t line items to avoid — they’re often functionally necessary — but they should be a deliberate spec, not an assumption.

What Inspection and Documentation Do You Require?

A part with no special inspection requirement gets checked against the print at final QC. A part requiring a full dimensional report, first article inspection, or material certification requires additional labor that’s separate from the machining itself.

How Fast Do You Need It?

Standard lead time lets a shop slot your job into existing machine capacity. Rush turnaround means bumping other jobs, potential overtime, and sometimes expedited material sourcing — all of which carry a premium, typically well above standard pricing.

The Real Question to Ask Before You Submit a Drawing

Every one of these factors is a lever, not a fixed cost. The fastest way to lower a quote isn’t to shop it around endlessly — it’s to look at the drawing and ask, for each tolerance, finish, and feature: does the part’s function actually require this, or is it a default?

FAQ

Does ordering more parts always lower the unit price? Usually, yes, because fixed setup cost spreads across more units — but the relationship isn’t linear, and at very high quantities other processes (like injection molding) may become more cost-effective than CNC. See our companion article on quantity pricing.

Is a tighter tolerance ever “free”? Occasionally, if the feature is small and the machine can hold it in the same pass as a looser tolerance elsewhere. But it’s the exception, not the rule.

Does material choice affect lead time as well as cost? Yes — less common alloys and sizes may need to be special-ordered, which adds time before machining even starts.

Can I get an accurate cost estimate before finalizing my drawing? Yes. Upload a CAD file to our instant quoting tool for a real-time price based on your actual geometry, material, and tolerances.

What’s the difference between standard and tight tolerance defaults on a quote? This varies by process and shop. We cover exactly what’s included in a “standard” quote versus a tightened one in Comparing CNC Quotes? Here’s What Tolerance and Finishing Line Items Actually Mean.

Sources: ASME Y14.5-2018, Dimensioning and Tolerancing (The American Society of Mechanical Engineers).

Related reading: GD&T Reference Guide · Comparing CNC Quotes: Tolerance and Finishing Line Items · Accurate Machine & Tool’s What Drives the Cost of Precision CNC Aerospace Parts? for a deeper look at cost drivers in regulated aerospace work.

Ready to see what your part actually costs? Upload your CAD file for an instant quote.

Most surprise CNC quotes trace back to four or five design choices, not the part itself — a tight tolerance on a feature that didn’t need it, a wall left too thin, a corner that should’ve had a radius. None of these make the part better. They just make it harder to cut.

Quick Summary

• Standard CNC tolerance is ±0.005″ — tightening it to ±0.002″ or below can add 25-50%+ to cost because it requires slower cuts, more inspection, and sometimes different equipment entirely.
• Walls thinner than 0.020″ risk breaking during machining or warping afterward, and should be avoided unless the design genuinely requires it.
• Sharp internal corners are physically impossible with a standard end mill — the tool is round, so every internal corner needs a radius, even if it’s a small one.
• Deep holes get expensive fast: past roughly 6x the hole’s diameter in depth, you often need specialized long drills or a different process entirely.
• The number of setups (how many times the part gets re-oriented in the machine) drives cost as much as the geometry itself — a part machinable from one side is cheaper than an identical part that needs to be flipped three times.

What Is Design for Manufacturability, Actually?

DFM is just the practice of designing a part so it’s cheap and reliable to make, not just correct on screen. A model can be dimensionally perfect in CAD and still be a nightmare to machine — the software doesn’t know or care how a cutting tool actually moves through material.

The gap between “this simulates fine” and “this machines fine” is where most quote surprises live. A tolerance that costs nothing to type into a CAD field can cost real money once it hits a machine, because achieving it might mean slower feed rates, extra inspection steps, or equipment your shop doesn’t normally reach for.

What’s the Standard CNC Tolerance, and When Do You Actually Need Tighter?

±0.005″ (±0.13mm) is the baseline most shops, including eMachineShop, work to without extra cost when a drawing doesn’t call out anything tighter. This default sits comfortably inside the broader tolerancing framework governed by ASME Y14.5, the standard most U.S. engineering drawings reference for how dimensions and tolerances are stated and interpreted in the first place. It’s tight enough for the overwhelming majority of mechanical parts — brackets, housings, spacers, mounting plates.

Tighter tolerances aren’t free. Dropping to ±0.002″ or below often means slower cuts to control tool deflection and heat, more frequent in-process measurement, and sometimes temperature-controlled environments to keep the material from moving as it warms and cools. The relationship isn’t a straight line either — manufacturing engineering research on tolerance allocation treats the cost of tightening a tolerance as a nonlinear function that climbs steeply once you’re past the easy range, not a modest step up. That’s not a shop being difficult — it’s the physics of holding a number that small.

The fix isn’t “never use tight tolerances.” It’s reserving them for the features that actually need them — a bearing bore, a press-fit diameter, a mating surface — and leaving everything else at the standard tolerance. A part with one tight-toleranced hole and everything else at ±0.005″ costs a lot less than the same part with every dimension needlessly tightened to match.

Why Do Thin Walls Cause Problems?

Below about 0.020″ (0.51mm), a wall gets fragile enough that it can crack or deform just from the force of the cutting tool passing by, before the part is even off the machine. If it survives that, it can still warp afterward as internal stress in the material releases.

If a thin wall is genuinely part of the design intent, it can usually still be made — but expect a conversation about it, and expect the finished part to vary slightly from the CAD model, since thin sections are the first thing to move. Where the wall thickness isn’t load-bearing or functionally required, thickening it removes the risk entirely and usually costs nothing.

Why Can’t CNC Mills Cut Sharp Internal Corners?

Because the tool is round. An end mill cutting a pocket or an internal corner leaves a radius behind, because a cylinder can’t produce a true 90-degree internal corner no matter how it’s moved. The radius left behind matches roughly the tool’s radius — a 1/4″ end mill leaves about a 1/8″ corner radius, minimum.

If a design calls for a genuinely sharp internal corner, the options are: add a small relief cut (a tiny extra notch past the corner) so the mating part can seat flush, accept the tool radius and design around it, or switch to a process like EDM that can produce a true sharp corner at added cost. Increasing a corner radius from 1mm to 3mm, where the design allows it, lets the shop use a larger, more efficient cutter and can meaningfully cut cycle time.

How Deep Can a Hole Actually Be Drilled?

Standard drilling gets impractical past roughly 6x the hole’s diameter in depth. Beyond that, the drill starts to wander off-axis, chip evacuation gets harder, and you often need a specialized long-flute drill or a switch to boring or gun-drilling, both of which cost more and take longer.

A 0.25″ diameter hole is fine down to about 1.5″ deep on standard tooling. Push that same diameter to 3″ deep and you’re into specialized-process territory. If the design allows it, a larger diameter hole reaches the same depth-to-diameter ratio without the extra process — worth checking before locking in a hole spec that seemed arbitrary in CAD.

Does the Number of Setups Actually Change the Price?

Yes, often more than the geometry does. Every time a part has to be unclamped, flipped, and re-fixtured to machine a different face, that’s added time, an added chance for misalignment, and in the worst case, a feature that ends up slightly out of position relative to a feature cut in a different setup.

A part with features on only one or two faces, oriented so the machine can reach them without repositioning, is meaningfully cheaper than a geometrically similar part with features scattered across four or five faces. This is also where multi-axis machines (4-axis, 5-axis) earn their cost premium — they can reach more of the part without a manual re-fixture, but the equipment itself costs considerably more to run, so it’s not automatically the cheaper option for a simple part.

Should You Specify GD&T, or Is a Basic Tolerance Good Enough?

For most parts, a basic linear tolerance is genuinely sufficient — GD&T is optional at eMachineShop and isn’t required to design or quote a part. GD&T earns its complexity on parts where a feature’s relationship to another feature matters more than its raw dimension — a hole that has to align with a mating hole across an assembly, a surface that has to stay parallel to a datum under load.

If you’re not sure whether a given feature needs GD&T, the practical test is this: does the part’s function depend on this feature’s relationship to something else, or just its size? Size alone, use a basic tolerance. Relationship to another feature or datum, that’s when GD&T starts paying for itself.

Frequently Asked Questions

Why did my quote come back higher than I expected? Almost always one of a handful of usual suspects: a tolerance tighter than the part needs, a wall thinner than the process likes, an internal corner sharper than a round tool can produce, or a hole deeper than standard tooling handles well. Checking a design against these before submitting it usually explains the number.

Does material choice affect these DFM rules? Yes — harder materials (stainless, titanium) amplify almost every issue on this list. A thin wall that’s marginal in aluminum is a real risk in stainless. A deep hole that’s routine in aluminum may need a different drill entirely in titanium.

Is DFM different for a one-off prototype versus a production run? The rules are the same, but the stakes shift. A single prototype might tolerate a slightly higher per-part cost from a DFM violation without much consequence. The same violation repeated across a thousand-unit production run compounds fast, which is why it’s worth fixing before scaling up, not after.

Can every feature that shows up fine in CAD actually be machined? No. CAD software will happily let you model an internal corner with zero radius or a wall thinner than any tool could leave standing — the software isn’t checking against what a cutting tool can physically do. That gap is exactly what a DFM review catches before it becomes a bad surprise on the shop floor.

What’s the single highest-leverage DFM fix? Reserving tight tolerances for the features that actually need them. It’s the one change that’s genuinely free to make (loosening an unnecessary tolerance costs nothing) and it’s the most common single line item driving up an otherwise ordinary quote.

---

Not sure if a design is fighting the process before you submit it? Upload the CAD file and we’ll flag anything worth a second look before it’s quoted.

Written by James Wright, eMachineShop

Two quotes for the same part can differ by a real margin and both be “correct” — because they’re not always quoting the same thing. Tolerance defaults vary by process, and a “standard finish” isn’t a fixed baseline across shops. Knowing what each term actually buys you is the difference between comparing apples to apples and picking the cheaper number without knowing why it’s cheaper.

Quick Summary

• 2D (sheet) parts default to ±0.01″ tolerance; 3D (machined) parts default to ±0.005″ — a real difference tied to the process, not the shop.
• At eMachineShop, detab and deburring are included in the job price, no extra charge — this includes whatever incidental brushing is needed to bring the part to 125 µin Ra or below.
• The dedicated Brushing finish — a distinct, uniform directional-grain finish — is a separate option with its own charge if you select it explicitly in the CAD software. That’s different from the incidental smoothing already included in standard processing.
• Some shops charge separately for detab, deburring, and hitting a controlled roughness value. eMachineShop bundles the first three into the standard price by default.
• If a part genuinely doesn’t need 125 Ra or better, the way to lower the price is to explicitly request as-cut — no detab, no deburring.

What’s the Difference Between 2D Tolerance and Mill Tolerance?

2D parts — sheet metal cut by laser, waterjet, or plasma — default to ±0.01″ unless a print specifies otherwise. 3D parts, machined on a mill or lathe, default to a tighter ±0.005″. This isn’t arbitrary: a cutting process moving through sheet stock has more inherent variation at the edge than a rotating tool removing material in a controlled cut, so the achievable baseline tolerance differs by process before a shop even factors in equipment or skill.

This matters directly for quote comparison. If one quote assumes ±0.01″ and another assumes ±0.005″ without either being stated explicitly, they’re not quotes for the same part — they’re quotes for two different implicit specifications, and the tighter one should reasonably cost more. Always check what tolerance a quote is actually built around before comparing the number at the bottom.

What Does “Standard Finish” Actually Include?

At eMachineShop, detab (separating the part from the stock it was cut from) and deburring (removing the sharp edge left behind) are included in every job’s price — no separate charge. That includes whatever incidental brushing is needed as part of deburring to bring the part to 125 µin Ra or below. Deburring and edge-finishing processes like these have been standard shop practice for as long as machining itself has been documented as a trade — this isn’t a new or exotic step, just one that’s easy to gloss over on a quote.

This matters more for some parts than others. Milled parts often land under 125 Ra just from the cutting process itself — a mill making a controlled pass through solid stock tends to leave a smoother surface than an edge cut by laser, waterjet, or plasma through sheet material. 2D (sheet-cut) parts are the ones that typically don’t get there on their own — as-cut sheet edges commonly sit well above 125 Ra, and getting under that threshold takes the deburring step that’s already included standard.

Worth distinguishing clearly: the dedicated Brushing finish — a separate, selectable option in eMachineShop’s CAD software that produces a uniform directional-grain surface — is not the same thing as the incidental brushing folded into standard deburring, and it carries its own additional charge if selected. If you don’t explicitly choose it, you’re not paying for it, and your part still gets the included detab-and-deburr processing regardless.

This is worth knowing specifically because surface roughness, commonly measured as Ra, is exactly the kind of spec that can quietly turn into a separate charge elsewhere. A quote that only covers cutting with no deburring included, and prices a controlled roughness value as an add-on, is quoting something different from a quote where deburring to 125 Ra or better is already part of the standard price. Comparing the bottom-line numbers without accounting for that difference isn’t really comparing the same part.

How Do You Get a Lower Price If You Don’t Need 125 Ra?

Ask for it explicitly. If a part is genuinely fine as-cut — an internal bracket nobody sees or handles, a component where a rough edge won’t affect fit or safety — request as-cut with no detab and no deburring. Skipping these steps entirely is a real, legitimate way to reduce cost on a part that doesn’t need them, but it has to be specified; it’s not the default you’ll get by leaving the finish field blank.

The tradeoff is straightforward: as-cut parts may still be attached to tabs from the stock sheet and will have sharp or rough edges from the cutting process, with no controlled roughness value at all. That’s a fine tradeoff for the right part. It’s the wrong one for anything that gets handled, needs to look clean, or has a mating surface where an untouched cut edge could cause a fit or seal problem.

Frequently Asked Questions

Does every part really come out at 125 Ra or below by default? Yes — detab and deburring, including whatever incidental brushing is needed, are included in the standard job price. If a part needs a tighter roughness spec than 125 Ra, that’s worth calling out explicitly since it goes beyond the standard.

Why does this matter more for 2D parts than milled parts? Because milled (3D) parts often land under 125 Ra just from the machining process itself, while 2D sheet-cut parts typically don’t — as-cut edges from laser, waterjet, or plasma commonly sit above that threshold and need the included deburring step to get under it.

Is the Brushing finish the same as the brushing included in standard deburring? No — the dedicated Brushing finish is a distinct, selectable option that produces a uniform directional-grain surface, and it has its own additional charge. The incidental brushing that’s part of standard deburring is included at no extra cost and isn’t the same finish.

Does tighter 2D tolerance ever match 3D machining tolerance? It can, on some features, but expect it to push the part toward a milling process instead of sheet cutting, since ±0.01″ is the realistic baseline for cut-sheet processes. A print calling for ±0.005″ on a nominally “2D” part is effectively asking for machined-part precision.

Does requesting as-cut mean sacrificing quality? Not if the part doesn’t need the roughness or edge condition that standard processing provides. It’s an informed cost tradeoff, not a shortcut — the part just won’t have tabs removed or edges deburred.

---

Have a part where standard finishing is more than you need? Upload your drawing and specify as-cut in your finish requirements — or check our RFQ tolerance defaults if you’re comparing quotes and want to confirm what’s actually being priced.

Written by James Wright, eMachineShop

If your part needs to survive wear at the surface but can’t afford to be brittle all the way through, case hardening is almost always the right call over through hardening — and picking the wrong one is one of the more expensive material mistakes we see on incoming drawings.

Quick Summary

• Case hardening builds a hard, wear-resistant layer on the surface of a part while leaving the core soft and tough. Through hardening hardens the entire cross-section uniformly.
• The two main case hardening methods are carburizing (adds carbon, needs a quench) and nitriding (adds nitrogen, doesn’t need a quench and causes far less distortion).
• Typical case depths run 0.010″–0.080″ depending on the process and soak time; surface hardness after carburizing commonly lands in the 58–62 HRC range.
• Low-carbon steels like 1018 and 8620 are built for case hardening — they physically can’t be through hardened because they don’t have enough carbon to form martensite all the way through.
• Case hardening is standard practice for gears, shafts, pins, and cams — anything that needs a wear surface but also needs to bend, flex, or absorb shock without cracking.

What Is Case Hardening?

Case hardening is a heat treatment that hardens only the outer layer, or “case,” of a part while the interior stays soft and ductile. The classic example is a gear tooth: the surface needs to resist wear from constant meshing, but the body of the gear needs enough give to absorb shock loading without snapping.

The process works by changing the surface chemistry of the metal — usually by diffusing carbon or nitrogen into it at high temperature — and then, in most (but not all) methods, quenching it to lock in a hard surface structure called martensite. The core, which never picked up the extra carbon or nitrogen, stays soft.

How Is Case Hardening Different From Through Hardening?

Through hardening changes the entire part uniformly. You heat the whole thing above its critical temperature, quench it, and the full cross-section hardens — assuming the steel has enough carbon to respond that way in the first place. A2 and D2 tool steel, 4340, and O1 are through-hardening steels; you’ll see them called out that way on a print because the hardness spec applies to the whole part, not a surface layer.

	Case Hardening	Through Hardening
What hardens	Surface only (a thin case)	Entire cross-section
Core condition	Stays soft and tough	Hard and brittle throughout
Typical steels	1018, 8620, 1010	4140, 4340, A2, D2, O1
Best for	Gears, shafts, pins — parts needing wear resistance + shock tolerance	Tooling, dies, structural parts needing uniform strength
Distortion risk	Lower (especially with nitriding)	Higher, particularly on large or thin sections

Neither one is “better” in general — they solve different problems. A gear tooth that’s through-hardened and brittle all the way to the core is a gear that cracks under shock load. A die that’s only case hardened won’t hold up to repeated deep impressions once the thin case wears through.

What Are the Main Case Hardening Methods?

Three show up on drawings most often, and they’re not interchangeable:

Carburizing diffuses carbon into the surface at 1,600–1,700°F, then quenches the part to form martensite. It’s the workhorse method — cheap, well understood, and it produces the deepest cases of the three. The tradeoff is distortion: the high heat and quench step can warp thin or asymmetric parts, which is why precision dowel holes and bores sometimes need to be reamed or ground after heat treat rather than before.

Nitriding diffuses nitrogen instead, at a much lower temperature (roughly 925–1050°F) and with no quench step at all. Because there’s no quench, distortion is minimal — parts can often be nitrided in their final machined state. The catch: nitriding needs alloying elements like chromium, aluminum, or vanadium to form the hard nitride compounds, so it doesn’t work on plain low-carbon steel the way carburizing does, and case depths run shallower.

Carbonitriding splits the difference — carbon and nitrogen both diffuse in, at a temperature between the other two. It’s often chosen for plain carbon steels that don’t have enough alloy content to nitride well but still need better hardenability than straight carburizing gives at a shallow case depth.

Method	Element Added	Needs Quench?	Typical Case Depth	Distortion
Carburizing	Carbon	Yes	0.010″–0.080″+	Higher
Nitriding	Nitrogen	No	0.002″–0.024″	Very low
Carbonitriding	Carbon + Nitrogen	Yes	0.005″–0.030″	Moderate

How Deep Does the Hardened Case Go?

Effective case depth is measured to a hardness boundary — commonly where the material drops to 50 HRC — and it’s controlled almost entirely by time and temperature at the furnace. Longer soak, deeper case. A light case for a small pin might run 0.010″; a deep case on a large gear tooth can push past 0.060″.

This is worth calling out on a print explicitly, because “case harden this part” without a depth spec leaves the heat treater guessing, and the resulting hardness profile can vary a lot between two shops interpreting the same vague callout differently.

Which Materials Can Be Case Hardened?

Low-carbon steels are the classic candidates — 1018 and 8620 especially, since their low carbon content means they physically can’t through harden but respond well to carburizing. We stock 1018 as a standard steel option and can source 8620 and other alloy grades on request; our steel materials page has the full rundown of what’s available by default.

Stainless steel can be case hardened too, but it’s a narrower conversation — austenitic grades like 304 and 316 don’t carburize the way plain carbon steel does, and low-nickel grades like 201/202 can pick up unwanted magnetism from cold work that complicates things further if magnetic permeability matters for the application. If you’re specifying stainless and need a hardened surface, it’s worth a direct conversation with us before locking in the material callout.

Does Case Hardening Cause Warping?

Carburizing can, particularly on thin sections or parts with big differences in cross-sectional thickness — the quench step is the culprit. Nitriding largely avoids this because there’s no quench at all. If a part has tight-tolerance features like dowel holes or a bore that has to stay round after heat treat, it’s common to leave those features slightly undersized before hardening and finish-ream or grind them afterward, rather than trying to hold final tolerance through the heat cycle.

What Parts Actually Need Case Hardening?

Gears are the textbook case, but the same logic applies to shafts, cam lobes, pins, and any wear surface that also has to flex or absorb impact. If a part’s failure mode under normal use is “the surface wears out,” case hardening is usually the answer. If the failure mode is “the whole part deforms or breaks,” through hardening or a tougher base material is the better fix.

Frequently Asked Questions

Can a case-hardened part be machined afterward? Not effectively on the hardened surface — the case is typically 58–62 HRC, well past what standard tooling handles cleanly. Any finish machining on critical features should happen before hardening, with grinding or reaming reserved for after, if needed.

Is case hardening the same as surface hardening? Yes, they’re used interchangeably. “Surface hardening” is the broader term; case hardening usually implies a diffusion-based method like carburizing or nitriding specifically, as opposed to induction or flame hardening, which harden the surface without changing its chemistry.

How is case depth actually measured? By cutting a cross-section, mounting and polishing it, and running a Vickers or Knoop microhardness traverse from the surface inward until the reading drops to the hardness boundary (commonly 50 HRC-equivalent). It’s a destructive test, so it’s usually done on a sacrificial sample rather than the actual part.

How much more does case hardening cost than through hardening? It depends more on part geometry and case depth than on the process itself. A shallow case on a simple pin is cheap; a deep case with tight distortion control on a complex gear can cost more than a straightforward through-harden, because of the extra masking, fixturing, and post-heat-treat finishing involved.

Why can’t 1018 be through hardened? Through hardening depends on carbon content high enough to form martensite across the full cross-section — roughly 0.3% carbon or more. 1018 sits at about 0.18%, which is enough to carburize a surface but not enough to harden all the way through, no matter how it’s quenched.

What’s the difference between case hardening and induction/flame hardening? Induction and flame hardening harden the surface by rapid localized heating and quenching — no carbon or nitrogen is added, so the base steel needs enough carbon already present to respond to heat treatment on its own. Case hardening changes the surface chemistry first, which is why it works on steels that induction hardening can’t touch.

Have a part that needs case hardening, and not sure which method or depth is right for it? Upload your drawing or CAD file and we’ll help you land on the right spec before it goes to the furnace.

Written by James Wright, eMachineShop