Tensile Stress in 3D-Printed Parts: Formula, Stress-Strain Data, Anisotropy, Test Methods and Design Limits

Tensile stress is the normal pulling force carried per unit of original cross sectional area. The engineering tensile stress formula is σ = F/A₀, and the tensile stress unit is usually MPa when force is in newtons and area is in mm². In 3D-printed parts, the calculation is straightforward; deciding what the result means is not. Print orientation, feedstock condition, specimen geometry, gauge length, strain measurement, tensile test speed, and the relationship between the load axis and layer plane can change the measured response. This guide explains how to calculate tensile stress, how to read tensile test data, and when a coupon result can or cannot support a part design.

EPOC CRAFTER’s 3D printing and additive manufacturing processes provide the manufacturing context for FDM, SLA, SLS, MJF, and metal AM parts discussed in this article.

1. What Is Tensile Stress in a 3D-Printed Part?

A practical tensile stress definition is force per unit area acting normal to a section while the part is being pulled. It is a stress state at a given load, not a material strength limit. In a standardized plastic tensile test, ISO 527-1 uses the original cross sectional area, so the reported quantity is engineering stress.

1.1 Tensile Stress Formula, Equation and Units

The tensile stress formula and tensile stress equation are the same engineering relation:

σ = F / A₀

where σ is engineering stress, F is tensile force, and A₀ is the original cross sectional area. For a rectangular tensile test specimen, A₀ = w₀ × t₀. This is also the normal stress formula for centered axial loading. In SI units, 1 N/mm² = 1 MPa. If you are asking what is tensile stress measured in, common stress units are Pa, MPa, or psi; N mm2 to MPa is numerically one-to-one when written as N/mm².

For a 3D-printed coupon, use the dimensional basis required by the tensile test standard and the actual measured specimen dimensions. Nominal CAD dimensions can differ from the printed section because of bead geometry, extrusion behavior, first-layer effects, or surface texture.

1.2 How to Calculate Tensile Stress: Case Example

The PA12-CF15 case used coupons with a mean width of 10.02 mm and thickness of 4.03 mm. The original area was 40.38 mm². At a tensile load of 1,200 N, the stress calculation is:

σ = 1,200 N / 40.38 mm² = 29.7 MPa

This force divided by area calculation gives the coupon’s engineering tensile stress. It does not prove that the actuator clevis bracket itself sees 29.7 MPa. The real part has an Ø8.0 mm pin hole, an 8.0 mm minimum ligament, contact at the pin, and a part-specific load path. Those features require part-level stress analysis.

Tensile stress vs compressive stress differs in the direction of the normal load. Tensile stress pulls the section apart, while compressive stress pushes it together. Tensile stress vs shear stress is a different comparison because shear acts tangentially to the section. This article stays with uniaxial tension.

2. Tensile Stress vs Tensile Strength, Yield Stress and Break Stress

Tensile stress vs tensile strength is the first distinction to make when reading a report. Tensile stress is the stress present at a stated load. Tensile strength is the maximum engineering stress reached during the tensile test. Yield stress and stress at break are separate points.

Quantity	Definition	Useful for	Do not infer
Applied tensile stress	F/A₀ at the current load	Current load state	Does not define material capacity
Yield stress	Stress at the applicable yield criterion	Onset or specified criterion of yielding	Not automatically the maximum stress
Ultimate tensile strength	Maximum engineering stress	Monotonic tensile strength comparison	Not a finished-part allowable stress
Stress at break	Engineering stress at fracture	Failure point under the test	Not automatically equal to UTS

2.1 Yield Strength vs Tensile Strength

The tensile strength vs yield strength distinction is functional. Yield strength or another specified yield criterion marks the onset of permanent deformation under that method; ultimate tensile strength marks the highest engineering stress. The same distinction applies when a search is phrased as yield strength vs tensile strength or tensile stress vs yield stress.

In the EPOC CRAFTER PA12-CF15 case, XY-parallel coupons reached 72.4 MPa mean maximum engineering stress and had a recorded 64.8 MPa value at 0.2% offset. Z-normal coupons reached 41.2 MPa and showed no distinct yield before fracture. The result is orientation-specific and should not be generalized to other builds.

The yield strength and 0.2% proof stress article covers offset criteria in detail; the present article only uses them to keep yield, UTS, and fracture from being mixed.

2.2 Engineering Stress vs True Stress

Engineering stress vs true stress differs by the area in the denominator. Engineering stress uses A₀. True stress uses the instantaneous area Ai. They can remain close at small uniform strain, then separate as the specimen changes cross section. A falling engineering stress after localized deformation does not prove that local true stress is falling.

Engineering stress: σe = F / A₀     True stress: σt = F / Ai

EPOC CRAFTER’s stress vs strain relationship in material testing and engineering stress strain curve interpretation cover the full engineering stress strain curve without duplicating that material here.

3. What a Stress-Strain Dataset Actually Tells You

A tensile test produces stress strain data, not just one tensile strength number. A useful polymer stress strain curve or plastic stress strain curve depends on how force, extension, and strain were measured. A stress strain diagram without that test basis can look precise while hiding a weak comparison.

3.1 Engineering Strain, Gauge Length and Extensometer

Engineering strain: ε = ΔL₀ / L₀

The strain formula and tensile strain formula use the original gauge length L₀. ISO 527-1 defines gauge length as the initial reference length in the specimen’s central region. An extensometer tensile test measures strain over that defined region. Crosshead displacement can include compliance in grips, the load train, and specimen regions outside the gauge length, so it is not automatically equivalent to specimen tensile strain.

In the case data, L₀ was 75.0 mm and strain measurement used a 50 mm clip-on axial extensometer meeting ISO 9513 Class 1. That information matters when comparing yield behavior, elongation at break, strain at break, or tensile modulus.

3.2 Tensile Modulus and the Small-Strain Region

Under ISO 527-1, tensile modulus is evaluated over the defined 0.05% to 0.25% engineering strain interval, using a chord or permitted regression approach. For a 75 mm gauge length, that interval spans only 0.150 mm of extension. Small strain errors therefore matter.

A high tensile modulus does not guarantee high polymer tensile strength, and a similar initial slope does not guarantee similar failure. The tensile modulus and elastic stiffness guide separates modulus from strength and part stiffness.

4. ASTM D638 vs ISO 527: What Must Match Before You Compare Data?

ASTM D638-22 and ISO 527 are both used for plastic tensile testing, but they are separate method systems. ASTM D638-22 is active. ISO 527-1:2019 remains the current published general-principles edition, although a revision is under development. ISO 527-2:2025 is the current edition for moulding and extrusion plastics and replaced ISO 527-2:2012.

Check	Why it changes comparability	What to request
Standard and edition	Methods and revisions are not interchangeable by name alone	Full designation and edition
Specimen geometry	Geometry affects area, strain field, and fracture	Specimen type, drawing, preparation route
Measured section	A₀ is the denominator in engineering stress	Actual w₀, t₀, and A₀
Gauge length and strain system	Strain, yield, and modulus depend on the measurement basis	L₀, extensometer type/class, crosshead use
Test speed	Polymers can be rate sensitive	Speed or strain-rate control by test stage
Conditioning and atmosphere	Moisture and temperature can change polymer response	Conditioning time, temperature, RH, test atmosphere
Direction and AM build state	Anisotropic 3D printing changes the load path	Load axis vs layer plane, raster, layer height, infill, walls
Replicates and statistics	A single specimen does not quantify scatter	n, individual results, mean, SD, rejected specimens

4.1 What ASTM D638 Does and Does Not Define for FDM

An ASTM D638 tensile test can support plastic tensile strength, yield or break-related stress, elongation, and tensile modulus under the stated method. ASTM D638 also recognizes anisotropic direction and reporting. It does not define universal FDM X, Y, or Z strength, raster angle, infill percentage, or an FDM anisotropy correction factor.

A report that says only ‘Z specimen’ is therefore weak. ‘Tensile axis normal to the layer plane’ is more useful because it describes the mechanical relationship. The same principle applies to FDM print orientation, print orientation tensile strength, and build orientation tensile strength.

4.2 Test Speed and Conditioning Are Part of the Result

The case used 1 mm/min through the modulus region and 5 mm/min to break. Coupons were conditioned to ISO 291:2008 23/50 for 88 hours and tested at 23.0 ± 1.0 °C and 50 ± 5% RH. Those are case conditions, not universal settings. A valid comparison reports them instead of assuming that every polymer tensile test uses the same tensile test procedure.

5. Why 3D Printing Tensile Strength Is Direction-Dependent

3D printing tensile strength is a property of a material plus a manufacturing state. In material extrusion, deposited roads and stacked layer interfaces create directional load paths. That is why 3D print orientation strength and FDM tensile strength cannot be reduced to a single polymer name.

5.1 Orientation and Failure Mode in the PA12-CF15 Case

Orientation	Tensile axis	Mean UTS	SD	Strain at break	Observed failure mode
XY-parallel	Parallel to layer plane	72.4 MPa	2.1 MPa	5.6%	Mixed bead rupture, limited inter-bead separation
Z-normal	Normal to layer plane	41.2 MPa	2.8 MPa	2.1%	Interlayer separation dominant

Source basis: EPOC CRAFTER case TS3DP-26-017, five specimens per orientation, same PA12-CF15 lot and nominal print settings.

The Z-normal mean was 43.1% lower than the XY-parallel mean in this test. That number is not a universal Z-direction reduction factor. It applies only to the stated feedstock lot, drying, machine, nozzle, 0.20 mm layers, ±45° raster, 100% infill, six perimeters, conditioning, and test method.

The failure evidence is as important as the percentage. XY coupons showed more bead rupture; Z coupons failed mainly along layer interfaces. That observation links FDM anisotropy and layer adhesion strength to an actual fracture mode instead of treating orientation as an abstract label.

5.2 Feedstock Identity and Lot Traceability

ISO/ASTM 52903-1:2020 addresses feedstock identity, composition, manufactured feedstock characteristics, packaging, lot identification, and quality assurance for material-extrusion plastics. It does not define the finished part’s tensile strength or replace a tensile test standard.

For traceable 3D printing mechanical properties, keep this chain intact: feedstock identity and condition, build parameters, specimen orientation, test condition, result. EPOC CRAFTER’s 3D printing material properties resource provides the broader material-selection context.

Whether the comparison involves PLA tensile strength, ABS tensile strength, PETG tensile strength, nylon tensile strength, plastic tensile strength, or polymer tensile strength, do not assume that a supplier datasheet and a printed coupon share the same manufacturing and test basis.

6. Can You Use a Datasheet Tensile Strength as an Allowable Stress?

No. A datasheet value can screen materials, but it is not automatically a material allowable stress or design stress for a finished 3D-printed part. ASTM D638 and ISO 527 define tensile testing, not a universal factor of safety.

Evidence level	What it represents	Useful decision	Does not prove by itself
Supplier datasheet	Supplier-defined material and test basis	Material screening	Printed-part capacity
Printed coupon	Defined feedstock, build, orientation, specimen and test	Process and orientation qualification	Local part stress
Finished part	Real geometry and load path	Part-level verification	Universal material property
Design allowable	Application-specific limit with required margin and evidence	Design release	Transfer to unrelated parts or conditions

6.1 Why the Coupon Calculation Does Not Prove the Bracket Is Safe

The case coupon gives 29.7 MPa at 1.20 kN because its measured A₀ is 40.38 mm². The actuator clevis bracket has a pin hole, a minimum ligament, contact loading, and possible bending. Dividing 72.4 MPa by 29.7 MPa and calling the result a factor of safety would compare a coupon UTS with a coupon-area stress, not the peak stress in the actual bracket.

A part-level review must reconnect material stress strain data to geometry, load path, service temperature, moisture, repeated loading, creep, fatigue, and the required reliability basis. Material properties for FEA can come from representative test data, but the material model and load case still have to match the behavior being simulated.

6.2 Procurement Check: Can I Compare These Two Tensile Numbers?

• Material identity: same polymer family, reinforcement, commercial grade, and traceable lot?
• Manufacturing state: same process family, orientation, raster, layer height, infill, walls, and relevant post-processing?
• Specimen basis: compatible geometry, actual cross section, gauge length, and preparation route?
• Test basis: same or compatible tensile test standard, edition, tensile test speed, atmosphere, and strain measurement?
• Statistics: replicated specimens with mean and scatter, not a single coupon?
• Failure mode: comparable fracture location and mechanism?
• Part relevance: does the finished geometry add holes, notches, thin ligaments, contact, bending, creep, fatigue, or temperature exposure?

When several answers are unknown, use the values for screening, not as equivalent design evidence. functional 3D printed prototype testing is the next evidence step when coupon data must be connected to a real assembly and load case.

7. EPOC CRAFTER Printed-Part Tensile Verification Case

Client C-274 / TS3DP-26-017 involved a PA12-CF15 actuator clevis bracket that transfers axial pull from a compact linear actuator to a pinned linkage. The design question was whether the part had enough tensile margin for a 1.20 kN service proof load and whether Z-normal loading was acceptable at the 8.0 mm minimum ligament beside an Ø8.0 mm pin hole.

7.1 Build and Test Basis

The feedstock was PA12-CF15 with 15 wt% chopped carbon fiber, lot PA12CF-260917-B. Filament diameter was 1.75 mm nominal and 1.74 to 1.76 mm measured. It was dried at 70 °C for 8 hours and transferred to a sealed dry box below 15% RH during printing.

FFF settings were a 0.40 mm hardened-steel nozzle, 0.20 mm layers, 270 °C nozzle, 90 °C plate, 55 °C chamber, alternating ±45° raster, rectilinear 100% infill, and six perimeters. Five specimens were tested per orientation.

The historical case used ISO 527-1:2019 and ISO 527-2:2012 Type 1A specimens, printed to shape. The test used an Instron 3367 universal testing machine, 5 kN Class 0.5 load cell, ISO 9513 Class 1 clip-on extensometer, 75.0 mm original gauge length, and the conditioning and speed values stated in Section 4. ISO 527-2:2012 is retained here because it is the actual test record; new work should use the current applicable edition.

7.2 Engineering Decision

The supplier datasheet listed 80 MPa tensile strength. XY-parallel printed coupons reached 72.4 MPa, or 90.5% of that value. Z-normal coupons reached 41.2 MPa, or 51.5%. The comparison showed that the same feedstock name did not produce one universal printed strength.

Production parts were therefore required to place the primary tensile axis in the layer plane. Z-normal loading was rejected for the critical feature, and the RFQ required matching lot and process coupons with each qualification build.

The case does not establish a universal PA12-CF15 allowable stress, fatigue life, creep limit, or XY-to-Z correction factor. It supports a process-specific orientation decision for this material lot, build condition, and test basis.

8. Engineering Questions From Current Search and Forum Demand

8.1 What Is Tensile Stress Measured In?

Tensile stress is measured in force per unit area. In SI engineering work, Pa and MPa are common; N/mm² is numerically equal to MPa. Tensile strain has no unit because it is a ratio of extension to original gauge length.

8.2 Which Tensile Test Standard Should I Use for a 3D-Printed Plastic?

Use the standard required by the governing material, customer, qualification plan, or test objective. ASTM D638 tensile test and ISO 527 tensile test methods can both be relevant to plastics, but neither creates a universal FDM build-orientation convention. State the exact standard and edition, specimen geometry, printing route, direction, conditioning, speed, and strain system.

8.3 Why Does Print Orientation Change Tensile Test Results?

Material extrusion creates deposited roads and layer interfaces. Loading parallel to the layer plane and loading normal to it can activate different load paths and fracture mechanisms. The PA12-CF15 case showed this directly, with 72.4 MPa XY-parallel and 41.2 MPa Z-normal mean UTS under the stated conditions.

8.4 Can ASTM D638 Tensile Strength Be Used Directly as an Allowable Stress?

No. ASTM D638 tensile strength is a coupon result under defined test conditions. It does not supply a universal factor of safety, fatigue limit, creep limit, or finished-part allowable stress. Use representative coupon data as one input, then evaluate the actual geometry, service load, environment, variability, and qualification requirement.

8.5 Why Can Published and Printed Tensile Data Differ?

A published value can come from a different specimen, material condition, process, orientation, speed, humidity, or strain system. Compare the test basis before comparing the number. This is why a tensile strength test should preserve enough metadata to explain the result rather than reporting MPa alone.