REPLACEMENT PART EVIDENCE
Measure a Replacement Part for 3D Printing After You Identify It
After part identification, measure the holes, shafts, clearances and tolerances a 3D printed replacement must match before you search libraries or commission a model.
What measurements matter for a 3D printed replacement?
SHORT ANSWER · CITABLE
Measure interfaces after you identify the part
Part identification tells you the product, part name and whether a catalog or standard component already exists. Measurement comes next: record the mating features that decide fit—hole centers, shaft diameter and shape, insertion depth, slot width, clip engagement, threads, wall thickness and the envelope the part must sit inside—plus clearance for movement and the load, heat or chemicals the part must tolerate. Use those numbers to search existing models or to brief a replacement-part request. If you still cannot name the part, start with the part identification for 3D printing guide before spending time on calipers.
The rest of this guide shows feature-by-feature methods, what to do when the broken part itself lies, an example measurement record, and how to check a downloaded model against your numbers.
Search the product before recreating the part.
Record the manufacturer, product name, exact model number, revision and every number molded or printed on the part. Search those identifiers with the part’s function before commissioning new geometry. A public replacement model, service drawing or compatible part may already exist. If you are not sure what the part is called or whether it is a standard component you can buy, start with the part identification guide; the replacement parts page explains how to tell a part number from mold, material and date codes.
Photograph the assembled product and the empty location where the part belongs. Context helps distinguish a cosmetic cover from a load-bearing latch and exposes interfaces that a loose fragment does not explain.
Measure the features that decide whether the part fits.
Overall length and width are rarely enough. Start with mating features: hole centers, shaft diameter and shape, insertion depth, slot width, clip engagement, thread, connector clearance and the surfaces that touch another object. Note the measurement units on every sketch.
Use calipers for small fit-critical features and a ruler for the overall envelope. Measure repeated features from a common datum instead of chaining one measurement to the next; chained measurements accumulate error. For hole spacing, measure between centers when possible. If the center is difficult to locate, measure outside-to-outside and subtract one hole diameter when the holes match.
Close the caliper gently and take every fit-critical reading at least three times, in different spots or orientations. If the readings spread by more than a few tenths of a millimeter, the part is oval, tapered or worn; record the spread instead of an average, because the spread is the information a designer needs.
- Overall envelope: maximum length, width and height.
- Interfaces: holes, shafts, threads, tabs, slots and contact faces.
- Clearance: space needed for movement, tools, cables or neighboring parts.
- Wall and feature thickness where strength or flexibility matters.
- Assembly direction and any feature that traps the part in place.
Feature-by-feature: how to measure, what to record, what usually goes wrong.
The same caliper gives different answers depending on how it meets the feature. Use the method in the second column, write down what the third column asks for, and check the last column before you send the numbers to a searcher or designer.
| Feature | How to measure | Record | Common replacement-part error |
|---|---|---|---|
| Round hole or bore | Inside jaws at the entry, at the bottom and again rotated 90°. On holes under about 5 mm, confirm with a drill-bit shank or pin that just slips in. | Smallest reading, depth, through or blind. | Small holes read undersize because caliper jaw tips are rounded and cannot reach the true diameter. |
| Shaft, pin or boss | Outside jaws at two heights and rotated 90°. Measure across flats on D-cut, hex or keyed shafts. | Largest reading, length, flat or key width. | A D-shaft recorded as a plain round shaft. The printed hole fits but spins instead of driving. |
| Hole spacing / bolt pattern | Measure the same edge of each hole (left edge to left edge) from one datum. On circular patterns, measure across opposite holes. | Every center from one datum, hole count, pattern diameter. | Chaining hole to hole. A 0.2 mm error per step becomes 0.8 mm across five holes. |
| Slot, groove or tab | Inside jaws for width at the mouth and at the root; depth rod for depth. | Width at both ends, depth, open or closed ends. | Molded slots taper (draft). Measuring only the mouth makes the replacement tab too thick at its root. |
| Snap clip or latch | Measure the hook overhang, arm length and arm thickness, then the ledge it locks behind on the mating part. | Overhang, arm thickness, insertion direction. | Copying a worn or whitened hook. The copy either fails to latch or breaks on first insertion. |
| Thread | Pitch gauge, or count crests over 10 mm; outside diameter with jaws; try a known screw or nut. | Metric or imperial size, pitch, length, internal or external. | Modeling small threads directly in FDM. Below roughly M5, plan a tapped hole, heat-set insert or purchased fastener. |
| Wall or rib thickness | Outside jaws, or depth rod against a flat, at several points away from the break. | Minimum and maximum thickness and where load passes. | Measuring next to a fracture, where the wall has already stretched, chipped or whitened. |
| Insertion depth and stop face | Depth rod from the mating surface to the stop. | How far the part enters and which face stops it. | Missing the stop face, so the part goes in too far or not at all. |
| Curved or freeform outline | Photograph square-on with a ruler in the same plane, or trace on paper; scan if available. | Scaled outline image plus three or more checked dimensions. | An angled photo distorts the outline; perspective error grows toward the image edges. |
A practical rule: for a feature the new part must go around (a shaft, a pin, a tube), keep the largest reading; for a feature the new part must go into (a hole, a slot, a socket), keep the smallest reading. The clearance is added later, as a separate decision.
When the broken part itself gives the wrong numbers.
Replacement parts are usually measured after they failed, so the object in your hand is often the least reliable evidence of its original size. The mating part—the hole, rail, ledge or screw it worked with—has usually changed less. Use this table to decide which surface to trust.
| What you see | Why the part misleads | Measure this instead |
|---|---|---|
| Polished, worn or whitened contact face | Material has been removed or stretched. | Measure the mating part: the hole the pin turned in, the rail the slider ran on. |
| Stripped or cracked screw boss | The hole has opened up and no longer shows the original size. | Measure the screw (diameter, length, head) and design the boss around the screw. |
| Warped or heat-distorted part | Flat references are no longer flat. | Measure the opening it sits in and an undamaged counterpart: the opposite side or a second unit. |
| Snapped part with missing fragments | The outline is incomplete. | Mirror the surviving left/right counterpart or measure the cavity; mark missing regions as unknown. |
| Aged rubber or soft plastic | Calipers compress it and age changes its size. | Measure the groove or seat. For seals, buy a standard O-ring by inner diameter and cross-section. |
| Reading just off a standard size | Measurement error around a nominal metric or imperial size (for example 7.9 mm, or 6.3 mm). | Check the nearest standard (8 mm; 1/4 in = 6.35 mm) against a known reference before rounding. |
Make a rough drawing that another person can audit.
A neat engineering drawing is not required. Draw separate top, front and side views, label the views, add arrows for dimensions and mark a consistent datum. Number the photos and point each measurement to a visible feature. State whether a value is measured, estimated or unknown.
Photograph the measuring tool in position for critical values. Keep the raw photos alongside the annotated images so a searcher or designer can check the interpretation. If a feature is inaccessible, measure the mating object instead and state the clearance you expect.
The free PrintWhat design brief provides a copy-ready structure for references, dimensions, interfaces, use conditions and deliverables.
Describe how precise—and how safe—the result must be.
Tolerance is not the number of decimal places in a measurement. It is the acceptable variation that still lets the part function. State what must slide, snap, rotate, seal or align. For an early test, it may be more honest to plan a fit-check prototype and adjust it than to promise a perfect first part from uncertain evidence.
Send the measured size, not the size plus your own allowance. If both you and the designer add clearance, the part ends up loose. Tell the designer what kind of fit each interface needs; the table below shows common starting clearances for desktop FDM prints in PLA or PETG, expressed as total difference in diameter or width. They are starting points for a test, not guarantees: printer calibration, material, orientation and hole size all move the result.
| Fit needed | Starting clearance (FDM) | What it should feel like |
|---|---|---|
| Press or friction fit | 0 to 0.1 mm | Pin holds without glue; may need a firm push. |
| Sliding or rotating fit | 0.2 to 0.4 mm | Moves freely with little wobble. |
| Loose or drop-in fit | 0.4 to 0.6 mm | Forgiving of warp and rough surfaces; visible play. |
Also record load, repeated motion, impact, sunlight, heat, moisture, chemicals and expected lifetime. These facts affect geometry and the eventual manufacturing process. PLA, for example, softens at temperatures a parked car interior or a dishwasher can reach. A model that looks correct is not proof that a material or print orientation is safe.
Check a downloaded model against your measurements before printing it.
When a search turns up a model that looks right, your measurement record becomes the test. Photos and titles cannot confirm fit; the interfaces can.
- Confirm units. Import the file into your slicer and compare the overall size with your envelope measurement. A model that is exactly 25.4 times too small or too large was exported in inches or read as inches.
- Measure the same interfaces on the model. Use the slicer’s or a free viewer’s measuring tool on hole spacing, shaft diameter and slot width, starting from the same datum you used on the real part.
- Classify each interface. Mark it as matches, needs adjustment, or wrong. One wrong mounting interface means a modification, not a straight print. If only a mesh is available, the STL vs STEP guide explains why small edits are harder.
- Print only the interface. Cut a thin slice—a few millimeters thick—that contains the mounting holes, clip or tabs, and test it on the product. It takes minutes and plastic instead of hours.
- Write down the result. Note material, orientation, which interface was tight or loose and by how much, so the next adjustment is measured rather than guessed.
Also confirm the license allows your intended use, and that the model is for the same product revision: left- and right-hand parts and mid-production revisions often look identical in photos.
Example: a measurement record for a broken battery cover.
The record below shows the format, using a TV-remote battery cover whose latch hook has snapped off. It is an illustrative layout, not data from a customer job; the point is the columns. Each row says where the value came from and whether it controls fit, so a searcher knows what to check and a designer knows what to test first.
| Item | What was recorded | Source | Role |
|---|---|---|---|
| Identity | Brand and model from the label inside the battery bay; part number molded on the cover | Read from label | Search these first |
| Envelope | Length, width and thickness of the cover | Measured, calipers | Sanity check only |
| Hinge tabs | Count, width, thickness; spacing measured edge to edge from the cover’s left edge | Measured, calipers | Fit-critical |
| Latch | Hook is broken: ledge depth and slot width measured on the remote body instead | Measured on mating part | Fit-critical |
| Tab angle | Cannot be measured with calipers | Estimated from a side photo | Confirm with a test slice |
| Use | Opened weekly, room temperature, no load | Stated by owner | PLA or PETG acceptable |
Notice that the broken latch is not measured at all. Its replacement is designed from the ledge on the remote body, and the one estimated value—the tab angle—is the first thing a test slice checks.
Six ways good measurements still produce a part that does not fit.
- Only the envelope was sent. Length, width and height let a designer draw a block, not place holes, clips or stops. Send interfaces first.
- A photo without scale. Size cannot be recovered from an image that lacks a ruler in the same plane, and an angled photo distorts proportions. Add at least one caliper value per view.
- Mixed units. Fractions of an inch beside millimeters on one sketch invite conversion errors. Choose one unit per drawing and write it on the drawing.
- The mating part was never measured. A replacement copied from a worn original inherits the wear. Measure the hole, ledge or rail it works against.
- Handedness or revision ignored. Mirror-image parts and revised housings share names and outlines. Photograph the part in place and record the product revision.
- Use conditions left out. A part can fit on day one and then creep, soften or crack from heat, load or chemicals. State the environment along with the dimensions.
Choose the cheapest route supported by the evidence.
Use the identity and dimensions to search existing model sources. If a licensed model matches most interfaces, define a bounded modification. Recreate the part only when search and adaptation cannot produce a defensible result. If you only have a photo of a similar product, the photo search guide shows how to turn it into searchable clues.
Keep model approval separate from production approval. The right manufacturing route may be FDM, resin, metal printing, machining, molding or a standard purchased replacement. PrintWhat starts with the required outcome and can route a request onward without forcing every object into one process.
SEARCH BEFORE REDESIGN
Turn the measurements into a replacement route.
Send the product identity, photos and critical dimensions. PrintWhat checks for an existing model first, then scopes a modification or reconstruction only when needed.
Start a replacement-part request