Measuring Metal 3D-Print Distortion After Build-Plate Separation

A metal part can change shape when it is separated from the plate on which it was printed. Measuring that movement is one way researchers study the effects of residual stress and strain, but the measurement must be tied to a particular geometry, material and procedure. NIST's additive manufacturing benchmark work illustrates why a single before-and-after photograph is not a complete distortion record.

Understand the question being measured

During metal additive manufacturing, rapid heating and cooling can leave residual stress and strain in a build. When material is removed or a part is separated from its support, the constraints on that part change and it may deflect. The size and direction of movement depend on the geometry and process; the fact that one test object moves does not establish the distortion of another design.

A 2023 NIST publication describes deflection measurements on IN718 nickel-based superalloy test objects prepared for an additive manufacturing modeling challenge. The authors measured the parts with a coordinate measuring machine after partial separation from the build plate. The work is a benchmark for research and model development, not a universal acceptance limit for production parts.

Keep the reference state visible

A deflection measurement compares positions under defined conditions. Record the part identifier, the point or surface measured, the reference used, and whether the measurement was made before or after a cut or other operation. If the reference surface itself moves, that movement can affect the reported result. An unlabelled distance in a spreadsheet is difficult to interpret without these details.

Photographs of the setup can supplement a diagram of the measurement points, provided they identify the same part and state. Note whether a surface was prepared before measurement and whether the instrument touched the part or used another method. These details help a later reviewer understand what was measured rather than assuming that two similarly named dimensions are identical.

The AMB2022-01 benchmark description explains that adjacent areas of the build plate were used as a reference in its distortion measurements to reduce the influence of plate warp. It also describes sampling points on the test object's ridge surfaces. Those details show how the reference and measurement locations belong to the result.

Document the separation sequence

For a production record, note whether a part remains attached to its plate, is partially cut, or has been fully removed when it is measured. Record the relevant support removal and post-processing steps in their actual order. A dimension measured while the part is constrained should not be silently substituted for a measurement taken after release.

Use the procedure required by the product's qualification plan and the equipment supplier. The NIST benchmark used a defined experimental geometry and sequence; it should not be copied blindly as a cutting instruction for another part. Cutting and handling metal builds can introduce hazards and must be carried out by qualified personnel with appropriate controls.

Distinguish geometry from stress

A coordinate measuring machine can provide dimensional observations at selected points. Those observations can help quantify a change in shape, but they are not automatically a direct measurement of the complete internal stress field. NIST's benchmark description lists several kinds of characterization, including residual strain and stress measurements as well as part distortion. Each method answers a different question.

If a report presents both a stress result and a dimensional result, identify the method, location, units and state of the part for each. A simulation can be compared with measurements only after the model's boundary conditions and the measurement procedure have been aligned. An attractive simulation image is not a substitute for evidence that its assumptions match the test.

Record the conditions needed for comparison

A useful build and measurement file includes the material and lot, geometry revision, machine, orientation, relevant build settings, support design and post-processing history. Add the measurement instrument, calibration information, points sampled, reference frame and uncertainty where available. The file should also identify the operator or laboratory and retain the underlying measurements.

When comparing two builds, avoid attributing a difference to a single setting if several conditions changed. A controlled experiment can help isolate variables, while ordinary production records may only support a more limited observation. If the source of a difference is unknown, state the uncertainty rather than presenting a suspected cause as a proven finding.

Use the benchmark as context, not a promise

NIST's work provides well-characterized test objects and measurements that researchers can use when developing or checking models. It does not guarantee that a commercial part will meet a particular tolerance or that a chosen heat treatment will be appropriate. Acceptance criteria should come from the actual product specification and an agreed inspection plan.

For a separate measurement topic, see our laser power check guide and browse the additive manufacturing archive. Keeping the part state, reference and measurement method together makes a distortion result much more useful.