Self-Centering Vise Repeatability Test

Short answer: test return after reclamping before judging finished parts. Freeze the machine, vise, mount, jaws, part, contact depth, clamping input, measuring method, and temperature state. Measure the same features. Fully release the part, remove and reseat it, clamp again, and record the raw result. Repeat the agreed series without correcting offsets. Then use a trial cut to test the complete process.

Repeatability is the spread in results when the same action is repeated under fixed conditions. Here, the action is loading and reclamping a part in one installed self-centering vise. Accuracy is different. Accuracy asks how close a result is to the required value. A setup can return to the same wrong place, or it can average near the target while returning with too much spread.

This guide tests an installed setup. It does not teach backlash adjustment, disassembly, or repair. If inspection finds damage, looseness, or unusual motion, stop. Use the selected model's controlled service procedure or ask the maker for written instructions before work continues.

Write the test question and acceptance limit first

“Is the vise precise?” is too broad for a useful test. Write one question that names the setup and decision. For example: “Does this installed vise return this part at the chosen datum within the shop's reclamp limit before a first-operation trial?”

A datum is the reference surface, axis, or feature used to locate and measure the part. Name it on the test sheet or marked drawing. Also name each measured feature, the direction of the reading, and the written acceptance limit. The limit must come from the drawing and process plan. Do not copy a catalogue number into a production release unless engineering has shown that it fits the full tolerance stack.

Agree on the cycle count before testing. There is no universal number. A small screening series may help expose an obvious problem, while a critical release may need more samples and a formal capability study. The process owner decides the plan from part risk, tolerance, and normal shop rules.

Before testing, decide how results will be judged. Record the minimum, maximum, and range for each feature. Range means maximum minus minimum. Do not remove an outlier without a written reason.

Freeze the installed setup so the test has one meaning

Create a setup ID. Record the machine make and model, machine state, vise model and serial number when available, mounting plate or adapter, fasteners, jaw set, jaw position, stops, supports, part identity, and material condition. Photograph the setup from useful angles.

Record how clamping input is applied. For a manual setup, note the approved tool and input method. For a powered setup, note the approved pressure or command state and how it was verified. Use only the selected model's instructions. A value from another vise family is not a substitute.

Also record:

• part orientation and which face meets each jaw;
• grip depth and any support under the part;
• contact surfaces that must be clean;
• machine warm-up state and recent cutting activity;
• room or machine temperature if heat can affect the decision;
• operator, measuring instrument, and instrument ID; and
• program and work-offset revision for the later trial cut.

Do not improve the setup halfway through the series. A jaw change, cleaned contact, new stop position, changed clamp input, or offset edit starts a new test. Close the first series and label the next one with a new setup ID.

Clean and inspect without turning the test into maintenance

Follow the machine maker's safe access and energy-control rules before hands enter the work zone. Bring the spindle and axes to the approved safe state. Use the shop's approved cleaning method and personal protection. Do not direct chips or fluid toward people, seals, or sensitive machine areas.

Clean the named seating and contact surfaces before the first cycle. Check for packed chips, burrs, damaged jaw faces, loose visible fasteners, or stock damage. Look at the mounting interface that can be inspected without removing the vise. This is a visual and functional precheck, not a teardown.

If a burr is part of the current normal condition, the process owner must decide whether to test that condition or correct it first. Do not quietly remove it and call the result a baseline. Record the choice. If the vise feels abnormal, a mount is loose, or safe grip is uncertain, stop. Do not keep cycling to collect a complete table.

Choose a stable part and map every contact

Use a stable gauge artifact or a representative production blank. A gauge artifact is a known item kept for checking a process. It does not have to be a laboratory block, but it must be clean, undamaged, and suitable for repeated seating. A thin or flexible part can add its own movement and make the vise harder to judge.

Mark part orientation. Identify the jaw contacts, stop contact, support points, and grip depth. Confirm that the part seats without rocking. If soft jaws are used, record their revision and condition. Do not change which end meets the stop between cycles.

Use the normal loading path every time. Include any hoist or other aid the job needs. A hand-loaded artifact may not represent a heavy or awkward production part.

Build a measurement plan the next shift can repeat

Choose an instrument that can resolve the movement that matters. A dial test indicator may suit an accessible surface. A machine probe may suit a feature inside the work zone. A coordinate measuring machine may be used for a removed or cut part. Each method measures a different chain, so state exactly what the result includes.

For every feature, write the instrument, contact point, approach direction, measuring force or probe routine when relevant, and zero method. Use the same machine position and approach. Avoid hand pressure on the part during a reading. If an instrument stand or stylus can move, include its check in the plan.

Take and record a reference reading after the first approved load. Do not make every later result look like zero on the display and lose the raw value. A signed change from the reference is easier to review. Note time and temperature when drift is possible.

Before cycling the part, repeat the measurement without releasing the clamp. This short check shows instrument and reading variation. If repeated readings move too much while the part stays clamped, improve the measurement method before judging the vise.

Run the clean-load-reclamp-measure series

A reclamp cycle means fully releasing the part, removing it from its contacts, loading it again by the fixed method, applying the fixed clamping input, and measuring the named features. Simply loosening and tightening without breaking contact may miss seating variation.

1. Confirm the setup ID, safe state, clean contact condition, and instrument check.
2. Load the marked part in the written orientation and bring it to the named contacts.
3. Clamp with the approved input. Do not add a hand tap unless the written method requires the same tap every time.
4. Measure every feature in the fixed order and record the raw signed values.
5. Fully release and remove the part. Record any chip, fluid, or seating event before cleaning.
6. Repeat until the preplanned cycle count is complete or a stop condition occurs.

Do not edit work offsets, jaw positions, stops, pressure, or operator method during the series. Do not stop after a good group of readings. If an event invalidates a cycle, keep the row, mark the reason, and follow the plan for whether a replacement cycle is needed.

Read spread, drift, and outliers separately

Spread is the distance between the high and low readings. Drift is a steady movement with cycle order or time. An outlier is one result far from the rest. These patterns point to different questions, but none proves a cause by itself.

Compare the range and any drift with the prewritten limit. Keep the full data table. An average alone can hide a large spread. If the series fails, change one controlled factor at a time and start a new setup ID. Swapping a known part, measuring method, or approved jaw set can help isolate the chain without dismantling the vise.

Add a controlled trial cut before production release

A passed no-cut test says the loaded part returned within the written reclamp limit for that setup. It does not prove the machining result. Cutting adds tool deflection, spindle condition, program path, heat, material response, clamp load, and inspection variation.

Use the approved program, tools, offsets, feeds, speeds, and inspection plan for the selected part. Confirm tool and rotary clearance before motion. Run the shop's normal dry-run, single-block, reduced-risk, or first-piece controls as required by the machine and process owner. This article does not replace those controls.

Choose trial features that can reveal movement in the directions that matter. Measure them with the released inspection plan. Keep no-cut reclamp readings and cut-part results as separate columns. If the no-cut result passes but the cut result fails, investigate the wider machining process. Do not “correct” the vise test with a work offset.

Release only the setup and part scope that passed

The release record should name the setup ID, machine, vise, mount, jaws, part family, clamping input, contact map, measurement plan, cycle count, raw readings, range, drift review, trial-cut result, limit, and approver. Attach photos and instrument records that your quality system requires.

Passing one part on one machine does not prove the same result on every machine, jaw set, grip depth, or part shape. Treat any material change as a new or revised test. Set a retest trigger for maintenance, jaw replacement, collision, mount removal, or another event your process owner considers important.

If the setup fails and the precheck finds wear, backlash, damage, or abnormal motion, close the test as failed or incomplete. Move the issue into the approved maintenance process. Do not mix repair work into the same data series and call the final rows a continuation.

What to send for a test-plan review

Send the machine make and model, selected vise model, mounting drawing, jaw and part drawings, material, grip depth, contact map, part tolerance, planned measuring method, and your draft acceptance limit. Include any existing setup data and state which machine and part the decision covers.

NEXTAS can review the selected hardware scope and identify missing interface or jaw information. Your process owner remains responsible for the sample plan, safe trial, measurement method, and production release. No catalogue image or remote review can certify the result on your machine.