The True Cost of a 3D-Printed Part: Beyond the Machine Hour

Most additive manufacturing quotations, and most internal business cases, start in the same place: how much material does the part consume, and how many hours will it occupy the machine?

Those are reasonable starting points. They are also where many AM cost estimates go wrong. For a metal part produced to an industrial standard, the hours a part spends in the printer may account for only a third of its cost. The rest is spread across activities that happen before the build starts and long after it finishes: build preparation, stress relief, plate removal, support removal, heat treatment, machining, finishing, inspection, scrap and the documentation that proves the part is what it claims to be.

This matters for two groups of people. Procurement teams comparing quotations need to understand what they are actually buying. Executives deciding whether to buy equipment need to understand which costs they are taking on. In both cases, the right question is not "What does the machine hour cost?" but "What does a good part, delivered and accepted, cost — and how does that compare with the best alternative?"

The US National Institute of Standards and Technology made a related point in its review of AM cost literature: in many instances, producing a product additively costs more than traditional methods, and the cases where AM is cost-effective tend to involve low volumes, complex geometry or customisation. That review is now more than a decade old and machine productivity has improved substantially since. The underlying structure of the cost, however, has changed less than the headline machine speeds suggest.

The cost stack

It helps to think of an AM part's cost in layers. The table below lists the main elements for a typical metal powder bed fusion part. Polymer processes share most of the same categories, usually at lower values and with shorter post-processing chains.

Layer Cost element What drives it

Capital Equipment acquisition and Purchase price, depreciation depreciation period, residual value

Facility                           Floor space, power, inert gas
                                                                      supply, powder handling, safety
                                                                      systems

Operating Machine utilisation Productive hours per year relative to available hours

Maintenance and consumables        Service contracts, filters,
                                                                      recoater blades, build plates,
                                                                      laser optics

Material                           Powder or feedstock                Price per kg, part volume,
                                                                      support volume, powder reuse
                                                                      limits and losses

Pre-build                          Design and build preparation       Orientation, support design,
                                                                      nesting, simulation, file
                                                                      preparation

Machine setup                      Powder loading, plate
                                                                      installation, chamber inerting,
                                                                      calibration checks

Post-build Stress relief Furnace time, often before removal from the plate

Plate and support removal Wire EDM or sawing, manual or machined support removal

Heat treatment / HIP Required properties; customer specification

Machining                          Critical surfaces, interfaces,
                                                                      tolerances AM cannot hold as-
                                                                      built

Surface finishing Blasting, polishing, chemical or abrasive flow finishing

Quality Inspection and NDT Dimensional inspection, CT

Layer Cost element What drives it

scanning, witness coupons, material testing

Scrap and failed builds Build failures, parts rejected at inspection

Qualification                        Process and part qualification       Material and process validation,
                                                                          first-article inspection,
                                                                          documentation

Each layer deserves a closer look.

Capital: depreciation only makes sense alongside utilisation

An industrial metal powder bed fusion system, with its powder handling, inert gas, furnace and safety infrastructure, represents a substantial investment. That cost is usually recovered through an hourly machine rate:

Machine rate per hour = (annual depreciation + annual fixed operating costs) ÷ productive hours per year

The denominator is where many business cases become optimistic. A machine that could theoretically run 8,000 hours a year will lose time to maintenance, calibration, material changes, build failures, set-up and — most importantly — a lack of work. A new in-house machine in a company still developing its applications may run productively for well under half its available hours. At that level, every part carries a substantial share of an idle asset.

Utilisation is not just a cost assumption. It reflects whether the organisation has enough qualified, repeatable demand to keep the machine busy. That is why the AM industry's own data increasingly emphasises utilisation and return. Wohlers Report 2026 described an industry facing higher expectations for utilisation and return, with services revenue growing far faster than equipment sales.

Material: the visible cost that is rarely the largest

Metal powders are expensive per kilogram compared with bar stock or castings, which makes material cost easy to notice. But for many parts it is not the dominant cost. Several factors determine actual material consumption:

• Part volume, which is often much lower than that of the equivalent machined billet. This is a genuine AM advantage for lightweight or lattice designs.

• Support structures, which consume material and then must be removed.

• Powder reuse. Unfused powder can usually be sieved and reused, but quality requirements may limit the number of reuse cycles or require blending with virgin powder.

• Losses during handling, sieving and cleaning.

For high-value alloys such as titanium or nickel superalloys, material can still be significant. For a part machined conventionally from an expensive alloy, where most of the billet ends up as chips, AM's

material efficiency can be a meaningful advantage. This is often described as the "buy-to-fly" ratio in aerospace.

Pre-build: engineering time that never appears on the machine timer

Before a build starts, someone must orient the part, design supports, nest multiple parts efficiently and, for critical parts, simulate distortion. For a new part this can take many hours of skilled engineering time. For a repeat part with an established build file, it may take very little.

This explains much of the difference between prototype and production pricing. The first build of a new part carries engineering effort that subsequent builds do not.

Nesting — the number of parts that can share one build — has a large effect on unit cost. Set-up, inerting and much of the preparation time are fixed per build, so they are divided among however many parts the build contains. A part that can be nested 20 to a plate behaves very differently from one that occupies the entire build volume alone.

Post-processing: where metal AM costs accumulate

For metal parts, post-processing is frequently the largest cost layer after machine time. A common sequence is:

• 1. Stress relief in a furnace, often while the part is still attached to the build plate to prevent distortion.

• 2. Removal from the plate, usually by wire EDM or band saw.

• 3. Support removal, which can be quick for well-designed parts and laborious for complex internal geometry.

• 4. Heat treatment or hot isostatic pressing (HIP) to achieve required mechanical properties and close internal porosity, where specified.

• 5. Machining of interfaces, sealing faces, threads and tight-tolerance features.

• 6. Surface finishing to meet roughness, cleanliness or cosmetic requirements.

Each step involves equipment, labour, handling and queue time. Machining deserves particular attention. AM parts are often marketed as "near-net shape," but any feature requiring a precision tolerance or sealing surface will usually be machined. That requires fixturing a complex organic shape, which is itself not trivial.

A practical test: if a printed part needs machining on most of its surfaces, the case for printing it should rest on something other than unit cost.

Inspection, scrap and yield

Quality costs scale with the criticality of the application. For a fixture or a non-critical bracket, a dimensional check may suffice. For a pressure-containing or flight-critical component, inspection may include CT scanning for internal defects, witness coupons built alongside the part and destructively tested, and full material certification.

Scrap enters in two ways:

• Build failures, where a build is lost to recoater crashes, distortion, delamination or equipment faults. The whole build's machine time, material and preparation are lost.

• Inspection rejections, where individual parts fail dimensional, surface or internal-quality checks after post-processing — by which point considerable cost has already been added.

The arithmetic of yield is unforgiving. If 92 percent of builds succeed and 97 percent of parts from successful builds pass inspection, overall yield is about 89 percent. Every good part must then carry the cost of roughly 0.12 failed parts. Mature, well-controlled production processes achieve better yields than this; new applications often achieve worse.

Qualification and documentation

For regulated or safety-critical industries, qualification can be the single largest cost of bringing an AM part into production, and it is almost always a one-time cost spread over the production volume.

Qualification may include establishing and freezing process parameters, characterising material properties, demonstrating repeatability across builds and machines, first-article inspection, and documentation for customer or regulatory approval. International standards increasingly set out frameworks for this. ISO/ASTM 52920 addresses qualification principles for industrial AM processes and production sites. ISO/ASTM 52904 addresses metal powder bed fusion processes intended for critical applications. Sector requirements in aerospace, energy and medical devices add further layers.

The commercial consequence is simple. Qualification cost per part is driven by lifetime volume. A qualification programme that is trivial over 10,000 parts can be prohibitive over 50.

Maintenance, consumables and the hidden overheads

Service contracts, replacement components, filters, build plates, inert gas, powder-handling equipment, operator training, health and safety systems and quality-system overhead all add to the annual fixed cost of running AM in-house. These rarely appear in the vendor's cost-per-part calculator, but they appear in the operating budget.

A worked example

The following example compares a metal AM part with a conventionally machined alternative. All values are invented for illustration. They are not market prices or quotations, and real costs vary widely by region, supplier, material, equipment and specification. The example shows how the cost stack behaves, not what any particular part should cost.

The part

An illustrative stainless steel bracket with moderate geometric complexity, around 0.6 kg finished mass. Expected demand is 200 parts per year for three years, a total of 600 parts. The part requires some machined interfaces and standard industrial inspection, plus a moderate qualification programme for the AM route.

Assumptions: in-house metal powder bed fusion

Item Illustrative assumption

Annual fixed cost of machine US$280,000 (depreciation, service contract, facility, gas and overhead)

Available hours / utilisation 8,000 hours at 65% → 5,200 productive hours

Machine rate ≈ US$54 per hour

Machine time per part (share of build) 5 hours

Material 0.6 kg + 25% supports + 10% losses, at US$90/kg

Parts per build 8

Build preparation / set-up per build 3 engineer hours at US$90; 2 technician hours at US$60

Post-processing per part                              Stress relief and heat treatment US$40; plate and
                                                      support removal US$60; machining 1.5 h at
                                                      US$110; finishing US$30

Inspection per part US$50 dimensional + CT sampling share US$50

Yield 92% build success × 97% inspection pass ≈ 89%

Qualification (one-time) US$60,000

Assumptions: conventional CNC machining

Item Illustrative assumption

Bar stock US$35

Machining 4.5 hours at US$110

Set-up and fixturing (share), deburring, finishing, US$140 inspection

Yield 98%

Fixtures and programming (one-time) US$8,000

Result

Cost element per good part AM, in-house (65% utilisation) CNC machining (illustrative, US$)

Machine time 269 495 (machining)

Material 74 35

Cost element per good part AM, in-house (65% utilisation) CNC machining (illustrative, US$)

Build preparation and set-up 49 incl. below

Stress relief / heat treatment 40 —

Plate and support removal 60 —

Machining of interfaces 165 —

Finishing 30 incl. below

Inspection 100 incl. below

Set-up, fixturing, deburr, finish, — 140 inspection

Subtotal before scrap 787 670

Allowance for scrap and failed 95 14 builds

Variable cost per good part 882 684

Qualification / one-time cost 100 13 per part (÷ 600)

Total cost per good part ≈ 982 ≈ 697

For this part, machining is cheaper by roughly 30 percent. Several features of the result are worth noting.

The printer is about a third of the cost. Machine time accounts for roughly US$269 of a US$982 part. Post-processing, inspection, scrap and qualification account for most of the rest.

Utilisation matters enormously. If the same machine runs at 35 percent utilisation instead of 65 percent, the machine rate rises to about US$100 per hour and the total cost per part rises to roughly US$1,240, about 78 percent more than machining.

Volume changes the one-time costs. The table below shows total cost per part across lifetime volumes, holding other assumptions constant.

Annual volume (3-year AM at 65% utilisation AM at 35% utilisation CNC machining (US$) life) (US$) (US$)

20 ≈ 1,880 ≈ 2,140 ≈ 820

50 ≈ 1,280 ≈ 1,540 ≈ 740

200 ≈ 980 ≈ 1,240 ≈ 700

1,000 ≈ 900 ≈ 1,160 ≈ 690

Annual volume (3-year AM at 65% utilisation AM at 35% utilisation CNC machining (US$) life) (US$) (US$)

At low volumes, AM's qualification cost dominates. At higher volumes, the gap narrows but does not close, because this part is not one AM is especially good at.

When the answer changes

Now change the part. Suppose the same function is currently delivered by an assembly of four machined components that are welded or brazed together and leak-tested, and the AM design consolidates it into a single part of similar printed cost. Illustratively:

• Four machined components at around US$230 each: US$920

• Welding or brazing, plus assembly labour: US$280

• Leak testing and inspection: US$120

• Assembly scrap of 4%, and US$25,000 of one-time welding fixtures and procedure qualification

The conventional assembly now costs roughly US$1,420 per unit at 600 lifetime units, against about US$980 for the AM part at 65 percent utilisation. The AM route is now around 30 percent cheaper, before counting any benefit from lower weight, fewer leak paths or a simpler bill of materials.

Change one more factor — lead time. If the conventional assembly takes ten weeks to source, and the part is needed to return a production line or field asset to service, each week of avoided downtime may be worth more than the entire part. Unit cost stops being the deciding factor.

This is the central lesson of AM economics: the same process, on the same machine, can be the expensive option or the cheap option depending on what it replaces.

How the answer changes across conditions

Factor Tends to favour AM Tends to favour conventional processes

Volume Low to moderate; many High, stable volumes able to variants amortise tooling

Geometry                           Internal channels, lattices,       Prismatic shapes, simple turned
                                   consolidated assemblies,           or milled parts
                                   organic shapes

Material Expensive alloys with poor Low-cost, easily machined machining buy-to-fly ratios materials

Tolerances and finish              Functional surfaces acceptable     Most surfaces need precision
                                   as-built or with limited           machining or fine finish
                                   machining

Qualification Non-critical parts, or high Critical parts at low lifetime lifetime volumes to spread

Factor Tends to favour AM Tends to favour conventional processes

qualification volume

Time value Downtime, obsolescence or Parts are stable, forecastable time-to-market is expensive and stocked

Outsourcing versus owning a machine

The worked example assumed in-house production. Many companies should start elsewhere.

Outsourcing to a service bureau or contract manufacturer converts fixed costs into variable ones. The provider spreads its machine cost across many customers, typically achieving higher utilisation than a new in-house user, and already has post-processing, inspection and quality systems in place. The customer pays a margin, but avoids capital commitment, staffing and the learning curve.

Owning equipment makes sense when:

• Demand is sufficient and stable enough to sustain high utilisation.

• The parts are strategically sensitive, or intellectual property concerns make outsourcing difficult.

• Speed of iteration between design and production creates competitive advantage.

• The company already has, or is prepared to build, post-processing and quality capability.

• AM is becoming a core manufacturing capability, not a side project.

A common progression is to outsource first, learn which parts generate repeatable value, and invest in equipment only when the demand profile justifies it. A sensible test before purchase is whether the company could fill at least half the machine's available hours with parts that are already qualified and ordered.

Inventory, supply chain and time-to-market effects

Some of AM's economic value does not appear in cost per part at all.

• Inventory. Printing on demand can reduce safety stock, obsolescence write-offs and warehouse costs for slow-moving spare parts.

• Minimum order quantities. AM avoids the minimum batch sizes tooling-based processes often impose.

• Supply chain resilience. A qualified digital file can, in principle, be produced by more than one qualified supplier or location.

• Time-to-market. For new products, removing tooling lead times can bring revenue forward by weeks or months.

• Engineering change. Design revisions do not strand tooling investment.

These benefits are real but should be quantified rather than assumed. A digital inventory strategy, for example, still requires that each part is qualified, that data is controlled and versioned, and that a capable producer is available when the part is needed.

How to read an AM quotation

For procurement teams, a few questions can reveal whether quotations are comparable:

• Does the price include stress relief, heat treatment or HIP where required?

• Which surfaces are machined, and to what tolerance?

• What inspection is included: dimensional only, CT, witness coupons, material certificates?

• Is the price for a first article, a prototype, or a qualified repeat production part?

• How many parts are assumed per build, and does the price depend on nesting with other customers' work?

• What happens to cost and lead time if a build fails?

• Who owns the build file, process parameters and qualification documentation?

Two quotations that differ by 40 percent may simply include different scopes.

Executive decision checklist

Before approving an AM part, programme or equipment investment, decision-makers can work through these questions:

• 1. Comparison. Have we compared total cost per good, accepted part against the best conventional alternative, not an outdated or poorly negotiated one?

• 2. Full stack. Does the estimate include build preparation, post-processing, machining, inspection, scrap and qualification, or only machine time and material?

• 3. Utilisation. If buying equipment, what utilisation does the business case assume, and what named, qualified demand supports it?

• 4. Volume. Over how many parts will qualification and engineering be spread?

• 5. Replacement. Is AM replacing a single part, or consolidating an assembly and removing joining, assembly and testing costs?

• 6. Time value. Have we put a financial value on lead-time reduction, downtime avoidance or time- to-market, and is that value real for this part?

• 7. Quality. Do we know the qualification path, and who will own it?

• 8. Make or buy. Would a qualified external supplier deliver the same benefit with less capital and risk at this stage?

• 9. Sensitivity. Does the case still hold if yield is five points lower, or utilisation 20 points lower, than planned?

• 10. Exit. If demand does not materialise, what happens to the equipment and the team?

Conclusion

The machine hour is a convenient number, which is why it dominates so many conversations about AM cost. But it describes only one layer of a production system that also includes engineering, post- processing, quality, yield and qualification.

The better question for any AM decision is the one procurement and finance teams already ask of every other process: what does a good, accepted part cost, delivered when it is needed, compared with the best realistic alternative? Sometimes the honest answer is that AM is the more expensive route.

Sometimes, once consolidation, lead time and downtime are counted, it is clearly the cheaper one. The value of a disciplined cost model is knowing which situation you are in before committing capital.

Omnidite Advisory supports technology evaluation and manufacturing investment decisions, including structured make-versus-buy and process-comparison analysis for additive and conventional routes.

This article provides general information and commercial commentary, not legal, investment, regulatory, medical or certification advice. Illustrative examples and calculations are not actual customer engagements or quotations.