Quoting Powder Bed Fusion Build Failures

Powder bed fusion (PBF) — whether laser or electron beam — is the workhorse of metal additive manufacturing. But every shop running DMLS, SLM, or EBM knows the sinking feeling when a 40-hour build crashes on hour 38. A recoater blade hits a curled edge. Supports detach. Gas flow shifts and porosity spikes. The whole build plate becomes expensive scrap.

Quoting these jobs means pricing risk that doesn’t exist in CNC or sheet metal. If you pad every quote with a flat “additive surcharge,” you lose competitive bids. If you ignore the risk, one failed build wipes out the margin on five good ones. The answer isn’t guessing — it’s modeling failure modes the same way you model machine time and material.

Identify the Failure Modes That Actually Cost You

Not all PBF failures are equal. Track your last 20–30 builds by failure type and you’ll typically see three categories dominate:

• Recoater crashes — curled edges, warped supports, or powder spatter stop the blade. Usually kills the whole build.
• Support failures — parts detach from the plate or supports fracture mid-build. Often recoverable if caught early, but adds labor and rework time.
• Porosity and density rejects — parts finish but fail CT scan or density specs. Hidden until post-process inspection.

Each has a different cost profile. Recoater crashes waste full machine time and powder. Support failures add labor but may save the build. Porosity rejects waste everything but catch late. Your quote should reflect which risks apply to this geometry.

Map Geometry to Failure Probability

Experienced operators know intuitively which parts are risky. Turn that intuition into a scoring checklist your estimators can use consistently:

• Large flat areas parallel to build plate → high curl risk → recoater crash probability
• Thin walls (< 1 mm) with tall aspect ratios → vibration during recoating → support failure
• Internal channels with no drain holes → trapped powder → post-process rejection
• Overhangs < 45° without self-supporting geometry → support dependency
• High thermal mass sections adjacent to thin features → residual stress → distortion

Assign each factor a weight based on your historical data. A part scoring high on recoater risk gets a different contingency than one scoring high on porosity risk.

Price Contingency by Failure Mode, Not a Flat Percentage

Replace “add 20% for AM risk” with line items tied to specific mitigations:

• Recoater crash contingency = (probability %) × (full build time + powder cost + plate prep). This is your “build insurance” line.
• Support rework allowance = estimated hours for manual support redesign + re-slice + partial rebuild time.
• Inspection fallback = CT scan cost + probability of reject × (rebuild cost).

Show these as optional line items or roll them into a “build risk adjustment” that the customer can see. Transparency builds trust — and lets technical buyers argue for design changes that lower the risk (and the price).

Build Orientation as a Quoting Lever

Orientation drives failure probability more than any other decision. But the “optimal” orientation for surface finish or support minimization may be the worst for recoater crashes. Quote multiple orientations with their risk profiles:

• Orientation A: minimal supports, 15% recoater crash risk, $X
• Orientation B: more supports, 3% recoater crash risk, $Y
• Orientation C: tilted 15°, balanced risk, $Z

Let the customer choose. Often they’ll pick the lower-risk orientation once they see the cost delta — and you’ve de-risked the build without eating margin.

Use Build Simulation Data If You Have It

Thermal-mechanical simulation (Netfabb, 3DXpert, Additive Works, etc.) predicts distortion, stress, and support loads. If you run sims pre-quote, feed the outputs into your risk model:

• Predicted max displacement > 0.5 mm → flag recoater crash risk
• Support reaction force > yield → flag support failure risk
• Residual stress concentration → flag distortion/porosity risk

Even a simplified simulation (thermal only) beats gut feel. If you don’t simulate every quote, run it on high-value or novel geometries and build a lookup table for similar parts.

Track Actuals to Calibrate the Model

The model only improves if you close the loop. For every PBF job, log:

• Quoted risk factors and contingency amounts
• Actual outcome: success, recoater crash, support rework, porosity reject
• Actual cost vs. quoted cost per failure mode

Review quarterly. Adjust probability weights. Remove contingencies that never trigger. Add new failure modes you missed. This turns quoting from a guessing game into a feedback-controlled process.

Communicate Risk Without Sounding Uncertain

Customers hate “it might fail.” They respect “we’ve modeled the three failure modes for this geometry, here’s the cost of each, and here’s how we mitigate them.” Frame the conversation around controlled risk:

• “We build a 10 mm sacrificial border on this geometry — adds 2% powder cost, reduces recoater crash risk from 18% to 3%.”
• “Orientation B adds 4 hours of support removal but eliminates the overhang that caused support failures on the last similar part.”

This shifts you from vendor to partner. It also justifies your price against shops that quote low and hope for the best.

Automate the Scoring in Your Quote Engine

If your quoting software can’t ingest geometry risk factors and output failure-mode contingencies, you’re doing this manually — which means inconsistently. A purpose-built quoting engine for digital manufacturing lets you encode the checklist, weightings, and contingency formulas once, then apply them automatically to every PBF RFQ. The estimator reviews, adjusts edge cases, and sends. Consistency scales; heroics don’t.

Solvi’s instant quoting engine is built for this — customizable risk models, geometry-driven pricing, and MES integration so actual build data feeds back into the next quote. See how it works.