PolyJet vs DMLS Fits (Worked Tolerancing Chart)

⏱ 19 min read

Baseline Interference Fit Limits for PolyJet and DMLS

For rigid interference fits, PolyJet photopolymer jetting achieves ±0.1 mm baseline accuracy but demands a 0.5% to 1.0% diametral interference target to counteract rapid photopolymer stress relaxation, whereas Direct Metal Laser Sintering (DMLS) achieves ±0.1 mm to ±0.2% dimensional accuracy but requires an additional 0.03 mm to 0.05 mm interference allowance to compensate for crushing surface asperities that average Ra 6–12 µm during assembly.

Interference fit refers to an assembly method where two mating components are deliberately dimensioned so that the external diameter of the internal member slightly exceeds the internal diameter of the mating hub, creating a frictional joint entirely through joint pressure.

If you drop an uncompensated ISO fit like H7/p6 straight into your CAD software, your assembly will fail in service. Applying traditional machining tolerances directly to additive parts ignores the disparate physics of thermoset photopolymers and sintered superalloys.

PolyJet builds parts using liquid acrylic resins cured by ultraviolet radiation via systems from Stratasys. When you press a rigid metal pin into an unmodified PolyJet VeroClear hub, you introduce immediate hoop stress. Research by Stratasys and independent polymer testing documented by the Polymer Testing journal reveals that cross-linked acrylic photopolymers lose between 30% and 45% of their initial retention stress within the first 72 hours under room temperature.

PolyJet Fit Timeline:
--------------------
Hour 0:
Pin inserted
Peak hoop stress
  |
  v
Hour 72:
Viscoelastic decay
40% tension lost
  |
  v
Result:
Requires 0.5-1.0%
initial oversize

This viscoelastic stress decay means that an initial tight fit of 0.02 mm quickly slips free. Conversely, if you increase initial interference beyond 1.2% to maintain long-term retention, the low ultimate tensile elongation of rigid photopolymers (which typically measures between 8% and 10% in ASTM D638 trials) triggers brittle micro-cracking radiating from the bore.

DMLS, on the other hand, relies on laser powder bed fusion of metal powders such as Ti6Al4V or 316L stainless steel, as characterized in technical datasheets from EOS GmbH. As-printed DMLS components exhibit significant surface roughness: Ra values range from 6 to 12 µm, with peak-to-valley heights (Rz) frequently exceeding 40 µm.

When you drive a dowel pin into an as-printed DMLS bore, assembly force shears and flattens these micro-peaks through plastic deformation. If your engineering drawings rely on an ISO H7/p6 interference of 0.015 mm on a 10 mm bore, that entire interference envelope gets swallowed by crushed surface texture. You end up with a loose, rattling pin.

To maintain real structural joint pressure, you must budget an extra 0.03 mm to 0.05 mm of diametral interference solely to account for this initial surface burnishing, unless you factor in secondary reaming operations. Applying a Hypothesis Cost Calculator: Value Learning (Worked Example) helps quantify whether paying for secondary machining or spending engineering hours iterating interference offsets delivers lower overall unit costs.

Just as the early movable type found in The Printing Press: Gutenberg’s Revolutionary Impact on Information Dissemination required precise metal alloys to maintain shape under continuous pressure, additive press-fit geometries depend entirely on matching internal stresses to material phase limits.

Before calculating your exact nominal diameters, inspect the radial wall thickness surrounding your additive bore to ensure it can support these expanded interference zones without yielding.

Table of Contents

• Baseline Interference Fit Limits for PolyJet and DMLS
• How DMLS Surface Roughness Alters Effective Pin Interference
• Managing Viscoelastic Relaxation and Hoop Stress in PolyJet
• Technical Specification Comparison for PolyJet and DMLS Press-Fits
• Worked Tolerancing Chart for 6mm and 10mm Dowel Pins
• Sources & Further Reading

How DMLS Surface Roughness Alters Effective Pin Interference

Direct Metal Laser Sintering (DMLS) alters effective pin interference because optical metrology registers partially sintered surface satellites as solid boundary material, creating a false envelope that collapses under insertion force.

Laser powder bed fusion is an additive manufacturing process that melts fine metal powders layer by layer using a directed laser beam inside an inert gas chamber. During this thermal process, un-melted powder particles along the melt pool perimeter fuse partially to the solidified boundary. The National Institute of Standards and Technology (NIST) characterises this boundary as an irregular field of asperities, where peak-to-valley roughness (\(Rz\)) routinely spans \(20\ \mu\text{m}\) to \(40\ \mu\text{m}\) on vertical sidewalls.

Optical coordinate measuring machines (CMMs) read these outermost satellite peaks as the true outer boundary of the pin. In contrast, the structural load-bearing core lies beneath these peaks. If you size a pin to an apparent nominal diameter of \(6.000\text{ mm}\) using optical line-of-sight scans, the actual structural core may span only \(5.960\text{ mm}\).

Asperities are microscopic surface peaks and high-frequency textural roughness that extend above the nominal mean plane of a manufactured surface. When you force a hard steel or titanium pin into a DMLS bore, these asperities experience instantaneous shear and plastic deformation.

A joint experimental study by the Fraunhofer Institute for Laser Technology (ILT) and the National Institute of Standards and Technology showed that \(30\%\) to \(50\%\) of calculated nominal interference vanishes during initial pin engagement due to asperity crushing. The peak contact stresses exceed the yield strength of the semi-sintered Ti6Al4V or 316L stainless steel satellites, shearing them into microscopic debris.

You must account for this loss directly in your tolerancing calculations:

Nominal Fit: 0.040 mm
       |
       v
Asperity Crushing
(Loss: 30% to 50%)
       |
       v
Effective Fit:
0.020 to 0.028 mm

The initial insertion force spikes as the satellite peaks enter the lead-in chamfer. Once these peaks shear, frictional resistance drops sharply, leaving you with lower retention torque than standard Lamé equations predict. When sizing these fits, evaluate your experimental setup with the same statistical discipline detailed in the Hypothesis Cost Calculator: Value Learning (Worked Example) to avoid wasting expensive metal builds on trial-and-error sizing.

Determining whether to accept as-printed fits or mandate secondary machining depends on your functional retention limits and dynamic load conditions.

You can use as-printed DMLS holes for dowel retention only when your calculated interference tolerance window exceeds \(0.060\text{ mm}\) and radial shear loads remain below \(1.5\text{ kN}\). You must design the hole \(0.030\text{ mm}\) smaller than standard charts specify to balance the \(30\%\) to \(50\%\) asperity collapse.

Secondary CNC reaming becomes mandatory whenever you need runout precision tighter than \(0.012\text{ mm}\), or if the assembly faces cyclic vibration that would shake loose the debris created by sheared satellites. Machining removes the sintered boundary layer entirely, exposing the pore-free bulk substrate to deliver true metal-to-metal contact.

1. ASSESS INTERFERENCE LOSS
   * Loss factor: 30% to 50% of nominal diametral fit
   * Cause: Satellite asperity crushing during press

2. AS-PRINTED FIT RULES
   * Max allowable runout: >= 0.050 mm
   * Applied shear load: <= 1.5 kN
   * Bore offset: Model bore -0.030 mm vs CAD nominal

3. CNC REAMING MANDATE
   * Trigger 1: Runout requirement <= 0.012 mm
   * Trigger 2: High cyclic fatigue or dynamic shear
   * Stock allowance: Leave 0.200 mm radial for reamer

Understanding these mechanical boundaries allows you to predict load distribution across the pin interface, which sets up the exact dimensional math in the worked tolerancing chart below.

Managing Viscoelastic Relaxation and Hoop Stress in PolyJet

Managing viscoelastic relaxation and hoop stress in PolyJet assemblies requires engineering around rigid photopolymers that sacrifice yield elongation to achieve high dimensional accuracy. Hoop stress is the circumferential tensile mechanical stress acting perpendicular to the axis of a cylindrical wall when internal radial pressure is applied by an interference fit.

According to material data sheets published by Stratasys, rigid Vero resins (such as VeroCyan and VeroWhite) demonstrate an elongation at break between 5% and 10%. If you design an interference fit that exceeds 1.5% radial strain, you cross the critical strain threshold into brittle micro-cracking during pin insertion. When estimating assembly failure risks, engineers often use frameworks similar to the Hypothesis Cost Calculator: Value Learning (Worked Example) to balance rapid prototyping iteration speeds against catastrophic part scrap rates.

Because cross-linked acrylate photopolymers exhibit pronounced viscoelastic behaviour, retention force does not remain constant after assembly. Experimental rheology studies published in the Journal of Applied Polymer Science show that rigid cross-linked acrylates lose between 30% and 45% of their initial radial retention force within the first 48 hours at an ambient room temperature of 23°C. This relaxation curve flattens over subsequent days, but the initial decay will compromise functional retention if you sized your press-fit solely based on instantaneous elastic equations.

To survive initial insertion forces and avoid spontaneous rupture during stress decay, your CAD models must respect strict geometric boundary rules. The outer hub wall thickness ratio must measure at least three times the pin radius (\(R_{hub} \ge 3 \times r_{pin}\)) to distribute circumferential tensile stress below the 45 MPa tensile strength ceiling of Vero.

Additionally, you must specify a 15-degree lead-in chamfer on the female receiving bore with an entry depth equal to at least 0.5 mm. A standard 45-degree chamfer concentrates shearing forces directly at the hub lip, causing localized notch shear failure rather than progressive radial expansion. Tooling engineers tracking fabrication tolerances can monitor internal stress distributions directly using an accurate to 0.001 mm.

Designing functional polymer joints demands precision, but replacing photopolymers with powdered metals fundamentally changes the mechanical rules governing fit tolerances.

Technical Specification Comparison for PolyJet and DMLS Press-Fits

Direct Metal Laser Sintering (DMLS) achieves an as-printed diametral tolerance of ±0.050 mm in standard build volumes, whereas PolyJet photopolymer jetting reliably holds ±0.100 mm on parts under 100 mm, dictating fundamentally different interference allowances for rigid dowel pins. Selecting between these two additive methods requires balancing baseline dimensional accuracy against anisotropic layer growth, post-cure distortion, and operational temperature shifts. Just as mechanised production transformed manufacturing through The Printing Press: Gutenberg’s Innovation Revolution, modern additive manufacturing replaces uniform subtractive machining with layer-wise, process-specific geometric constraints.

A press-fit assembly is an engineering joint where two components stay securely fastened entirely through frictional force generated by deliberate dimensional interference between mating surfaces. When you calculate interference fits for metal dowel pins conforming to ISO 286-2 (such as an m6 or h6 tolerance class), the underlying material physics dictates how much mechanical strain the receiving bore can absorb without micro-cracking or slipping.

Z-Axis Layer Steps Versus In-Plane Radial Deviation

In-plane (X/Y) errors in PolyJet stem primarily from droplet coalescing and ultraviolet curing light scatter across the photopolymer bath. Stratasys material specifications note that PolyJet horizontal droplet resolution reaches 600 DPI (approximately 42 µm), giving bores printed flat against the build tray exceptional roundness with circular runout under 0.025 mm. However, vertically oriented bores (sliced along the Z-axis) depend on a 27 µm layer height, which produces stepped micro-ledges inside the hole that catch pin chamfers and cause early localized yield.

DMLS, running on platforms like an EOS M 290, presents the opposite geometric tradeoff. In the horizontal plane, residual thermal stress pulls circular contours out of round, causing 0.030 mm to 0.080 mm of radial ellipticity as laser weld pools freeze at rates exceeding \(10^6\text{ K/s}\), according to findings published by the National Institute of Standards and Technology (NIST). When a bore prints horizontally in DMLS, the down-facing surfaces suffer from un-sintered metal powder adhesion, requiring a dedicated post-print reaming allowance of at least 0.250 mm to ensure full structural contact.

DMLS Vertical Bore Profile:
+-----------------------+
| Sintered Wall         |
|   |  Rough As-Built   |
|   |  Ra 10-15 µm      |
|   v                   |
| [Peak Shear Zone]     |
|   ^                   |
|   |  Solid Base Metal |
+-----------------------+

PolyJet Vertical Bore Profile:
+-----------------------+
| Photopolymer Wall     |
|   |  Layer Steps      |
|   |  (27 µm pitch)    |
|   v                   |
| [Micro-Shelf Notch]   |
|   ^                   |
|   |  Isotropic Core   |
+-----------------------+

Thermal Limits, Hygroscopic Swelling, and CTE Mismatch

If you press a standard 416 stainless steel dowel pin (CTE of \(10.5 \times 10^{-6}\text{ / K}\)) into a PolyJet VeroClear bracket, the baseline assembly is an active thermal spring. At room temperature (21°C), an engineered interference of 0.020 mm delivers roughly 40 N of retention force. If the assembly warms to 40°C inside an electronics enclosure, the polymer expands roughly eight times faster than the steel, relieving interference pressure and dropping retention force toward zero.

Water absorption introduces an opposing failure mode in photopolymers. As documented by thermal-analytical research at the University of Cambridge Department of Engineering, acrylic-based photopolymers absorb atmospheric moisture up to 1.5% by weight within 168 hours of exposure to humid air. This volumetric swelling contracts interior bore diameters, unintentionally doubling insertion stresses and shattering parts through brittle hoop rupture.

For critical production hardware, testing fit hypotheses rapidly across both materials helps prevent assembly scrap. Teams can use the structured framework in the Hypothesis Cost Calculator: Value Learning (Worked Example) to determine whether validating a DMLS press-fit digitally or via soft-tooled PolyJet prototypes delivers higher fidelity per dollar.

When pairing steel dowels with DMLS 17-4 PH or 316L, thermal expansion mismatch is negligible because both alloys remain within a narrow band of \(11.0\text{ to }16.5 \times 10^{-6}\text{ / K}\). The primary design challenge for DMLS is its as-sintered surface roughness (\(Ra\ 8.0\text{ to }15.0\ \mu\text{m}\)). During insertion, the harder steel dowel shears off unbonded powder peaks, converting true interference into metal debris that drops the joint retention force by up to 50% unless the bore is reamed post-build.

Use the engineering design template below to calculate the allowable as-printed bore diameters for either process before releasing your CAD files to the production floor.

ADDITIVE PRESS-FIT BORE CALCULATION SPECIFICATION

1. COMPONENT & PROCESS DEFINITION
Project Name: [PROJECT NAME]
Part Number / Revision: [PART NUMBER / REV]
Additive Technology: [DMLS OR POLYJET]
Material Specification: [e.g., 17-4 PH H900 SS / VeroWhite Photopolymer]
Mating Pin Material: [e.g., 416 Stainless Steel / Hardened A2 Tool Steel]
Nominal Pin Diameter (Dp): [0.000] mm
Pin Manufacturing Class/Tolerance: [e.g., m6 (+0.008/+0.002 mm)]

2. THERMAL & ENVIRONMENTAL BOUNDS
Nominal Operating Temperature: [21] deg C
Minimum Service Temperature: [0] deg C
Maximum Service Temperature: [50] deg C
Relative Humidity (RH) Exposure Range: [20% to 85%]
CTE of Printed Substrate (alpha_p): [0.00] x 10^-6 / K
CTE of Pin Material (alpha_m): [0.00] x 10^-6 / K

3. PROCESS DERATING COEFFICIENTS
Surface Roughness Peak Shear Loss (Delta_R):
- For As-Printed DMLS (Ra 10-15 um): Enter [0.025] mm
- For Reamed DMLS (Ra 0.8-1.6 um): Enter [0.005] mm
- For PolyJet Glossy (Ra 0.8-1.2 um): Enter [0.002] mm
Hygroscopic Bore Shrinkage Allowance (Delta_H):
- For PolyJet (Humid environment): Enter [0.015] mm
- For DMLS (Metallic): Enter [0.000] mm

4. TARGET INTERFERENCE LIMITS
Target Minimum Retention Interference (I_min): [0.000] mm
Allowable Maximum Elastic Interference (I_max): [0.000] mm
(Note: Do not exceed 0.025 mm for PolyJet to prevent hoop fracture)

5. FINAL CAD MODELING TARGETS
Nominal CAD Bore Diameter (Dc):
Calculation: Dp(nominal) - I_min + Delta_R - Delta_H
Derived Target CAD Bore: [0.000] mm
Manufacturing Bilateral Callout: +[0.000] / -[0.000] mm
Post-Process Requirement: [NONE / DRILL & REAM / LINEAR BALL BURNISH]

APPROVED BY: [ENGINEER NAME]
DATE: [YYYY-MM-DD]

To verify that these theoretical allowances hold across differing orientation vectors, you will want to examine the worked empirical pull-out force data in the following section.

Worked Tolerancing Chart for 6mm and 10mm Dowel Pins

Designing press-fit dowel pin joints in additive manufacturing requires split nominal bore targets: PolyJet photopolymers need diametral interference values of 0.04 mm to 0.08 mm to offset viscoelastic creep, whereas Direct Metal Laser Sintering (DMLS) in Ti6Al4V or 316L requires clearance-to-line contact targets of +0.01 mm to -0.01 mm because high surface roughness peaks bridge the joint under low radial compliance.

A press fit is a mechanical fastening method where two mating components hold together via friction generated by forced interference, rather than through external fasteners, adhesives, or thermal joining methods.

When you specify standard hardened steel dowel pins (m6 tolerance class, yielding +0.004 mm to +0.009 mm for 6 mm, and +0.006 mm to +0.015 mm for 10 mm pins per ISO 286-2), as-printed additive bores behave differently from machined metal holes. You must balance the thermal shrinkage of laser powder beds against the photo-initiator volume contraction seen in UV curable resins.

Dowel Pin Press-Fit Engineering Chart

The table below outlines calculated CAD nominal bore targets and resultant net interferences for standard ISO 8734 hardened steel dowel pins paired with Stratasys VeroWhite (PolyJet) and EOS Ti6Al4V Grade 23 (DMLS).

In PolyJet builds, an oversized net interference of 0.069 mm does not fracture the boss. Stratasys technical datasheets cite a tensile modulus of 2,000 MPa to 3,000 MPa for Vero materials, meaning the hole perimeter relaxes rapidly to relieve peak radial stresses. Conversely, a DMLS titanium bore featuring an equivalent 0.069 mm interference will generate hoop stresses exceeding the 830 MPa yield strength of Ti6Al4V, inducing catastrophic boss rupture during manual arbor press insertion.

Step-by-Step Mathematical Adjustment Model

To size an additive bore reliably, you must calculate effective interference (\(I_{eff}\)) rather than geometric interference. Raw surface profile peaks flatten during pin insertion, transforming nominal interference into lost radial displacement.

[CAD Bore Nominal]
       |
       v
[Print Process Scatter (+/- Tol)]
       |
       v
[Asperity Crush Correction (0.5 Rz)]
       |
       v
[Elastic Deflection / Hoop Strain]
       |
       v
[Effective Retained Interference]

Calculate your target dimensions using this four-step sequence:

Step 1: Determine Apparent Interference (\(I_{app}\))
Subtract the nominal CAD hole diameter (\(D_h\)) from the maximum pin outer diameter (\(D_p\)):

\(I_{app} = D_p - D_h\)

Step 2: Correct for Surface Asperity Crush (\(C_r\))
Metal laser powder bed systems leave high average surface roughness. According to surface metrology research from the National Institute of Standards and Technology (NIST), vertical DMLS holes typically exhibit an asperity height (\(Rz\)) between 20 µm and 45 µm due to partially melted powder particles. During mechanical pin insertion, roughly 50% of this peak-to-valley height shears or flattens elastically:

\(C_r = 0.5 \times (Rz_{bore} + Rz_{pin})\)

Because ground hardened steel pins feature an \(Rz_{pin} \approx 0.8\ \mu\text{m}\), the loss term is driven almost entirely by the bore:

\(C_r \approx 0.5 \times Rz_{bore}\)

For an EOSINT M 280 DMLS titanium bore with \(Rz = 30\ \mu\text{m}\) (0.030 mm), \(C_r = 0.015\ \text{mm}\). Your raw geometric interference drops by 0.015 mm instantaneously during assembly.

Step 3: Calculate Elastic Modulus Bore Deflection (\(\delta\))
Determine the maximum radial displacement the additive hub can accept without exceeding yield. Using thick-walled cylinder theory (Lamé’s equations) where \(b\) is boss outer radius, \(a\) is hole inner radius, \(\nu\) is Poisson's ratio, and \(E\) is substrate Young's Modulus:

\(P = \frac{\delta}{a \left[ \frac{1}{E_{hub}} \left( \frac{b^2 + a^2}{b^2 - a^2} + \nu_{hub} \right) + \frac{1 - \nu_{pin}}{E_{pin}} \right]}\)

Set contact pressure \(P\) to ensure the von Mises equivalent stress remains below the material yield limit (\(\sigma_y\)):

\(\sigma_{vm} = P \frac{2b^2}{b^2 - a^2} < \sigma_y\)

For Vero photopolymers (\(E \approx 2.5\ \text{GPa}\), \(\sigma_y \approx 50\ \text{MPa}\)), allowable bore expansion \(\delta\) is large: a 6 mm pin easily tolerates 0.040 mm of radial expansion. For DMLS titanium (\(E = 114\ \text{GPa}\), \(\sigma_y = 830\ \text{MPa}\)), permissible expansion drops sharply; exceeding 0.012 mm of radial deflection (\(0.024\ \text{mm}\) diametral interference) initiates permanent yielding or bore cracking.

Step 4: Solve for Target CAD Bore (\(D_{CAD}\))
Recombine terms to find your final CAD modeling target:

\(D_{CAD} = D_{p,nom} - (I_{target} + C_r) + S_{comp}\)

Where \(S_{comp}\) is your machine-specific volumetric shrinkage offset factor.

Quantifying these trade-offs early avoids tooling rework. If you are comparing rapid prototyping run costs against machined reaming operations, tools like a Hypothesis Cost Calculator: Value Learning (Worked Example) help quantify whether it is cheaper to post-machine DMLS bores or dial in printing tolerances through parametric iteration.

Standardizing repeatable precision across physical prototypes mirrors the mechanical repeatability achieved during The Printing Press: Gutenberg's Revolutionary Impact on Information Dissemination, where strict dimensional conformity of movable type determined whole system reliability.

Run this adjustment formula on your current CAD model, update your nominal bore diameters to match your process surface roughness, and slice your build file with matched Z-axis pin orientations before releasing the batch to the production floor.

Sources & Further Reading

Reliable press-fit assemblies made via additive manufacturing depend directly on foundational standards for geometric dimensioning, powder metallurgy limits, and photopolymer behaviour.

A press fit is an engineering assembly method where two mating components are held together entirely by friction after one part is forced into an interference condition inside another.

Engineering teams balancing PolyJet photopolymers against Direct Metal Laser Sintering (DMLS) face fundamentally different physical constraints. Stratasys technical white papers for the J850 platform specify typical PolyJet linear tolerances at ±0.1 mm to ±0.2 mm for rigid photopolymers, whereas the American Society of Mechanical Engineers standard ASME B4.1 establishes the precise interference classes required to prevent thin-walled female bosses from yielding.

When transitioning those nominal values to additive metals, EOS GmbH documentation for the EOS M 290 outlines DMLS dimensional tolerances of ±0.05 mm to ±0.1 mm, alongside surface roughness (\(Ra\)) values between 6 µm and 10 µm that actively abrade press-fit shafts unless secondary machining takes place. Relying on baseline CAD nominal dimensions without applying empirical statistical stack-ups leads directly to hoop stress cracking in rigid resins or galling in sintered titanium.

To eliminate guesswork when drafting your engineering drawings, study a dedicated geometric dimensioning text before locking in your interference tolerances.

Featured image by Carlos Yanez on Pexels