Kontakt- och beröringsfria system är mätvägar, inte automatiska kvalitetsrankningar. Rätt val börjar med den kontrollerade egenskapen och delens skick: vilken egenskap definierar rakhet, hur delen stöds, om kontakt är tillåten, hur snabbt det rör sig, vilka miljöeffekter som finns och hur resultatet kommer att korreleras med acceptans.
*Engineering koncept illustration. Själva sonden, stöd och acceptansberäkning kräver en kontrollerad mätmetod.*
What Each Measurement Family Physically Does
Every straightness gauge, whatever its brand, answers one question: where is the surface, referenced to what, under which part state? Contact systems answer it by resting a mechanical tip on the surface and converting tip displacement into an electrical signal — the principle behind the LVDT multipoint measurement arrangements used on straightening machines. Non-contact systems answer it without touching: laser displacement sensors and light-curtain arrays triangulate or shadow the surface, and vision systems resolve edges from images. Neither family measures “rakhet” directly — both produce surface positions that a calculation then turns into a straightness value by choosing a datum, a filter and a reference line.
That last point is where most comparison discussions go wrong. The sensor is one link in a chain: part presentation, datum sittplats, stödvillkor, sampling positions along the length, the evaluation algorithm, and the acceptance rule all sit between the raw reading and the pass/fail decision. Two systems with identical sensor specifications can disagree completely if one samples the part at five stations and the other at fifty, or if one references end centers and the other references a best-fit line. Method comparison, not sensor comparison, is the real task.
Kontakta Mätning: Strengths and Failure Modes
Contact measurement remains the default on production straightening equipment for concrete reasons. The measurement is referenced to a physical support geometry, so the datum is explicit and mechanically repeatable. The tip integrates over a small area rather than reading individual surface texture peaks, which gives stable readings on ground or turned surfaces. The hardware is robust to the coolant, chips and ambient light found next to a press, and calibration against a physical reference — a gauge block, a master part — is straightforward and auditable.
Its failure modes are equally concrete. Tip force leaves witness marks on soft, pläterad, polished or coated surfaces — the reason contact strategy is inseparable from praxis för ytskyddsverktyg. A worn or flattened tip changes the effective integration area and biases readings slowly, which is why tip condition belongs in the calibration schedule, not in the breakdown list. On rotating parts, eccentricity of the centers or supports couples into the reading as a once-per-rev signal that can be mistaken for bend if the setup is not diagnosed first — the distinction worked through in rundhet vs böj vid roterande mätning. And throughput has a mechanical ceiling: a tip must track the surface, so speed limits come from gapping and stylus dynamics rather than from electronics.
Where contact fits best
Contact methods fit stiff parts with machined surfaces, tight and explicit datum schemes, and acceptance rules written against a mechanical gauge. They are the natural choice when the production machine must measure, correct and re-measure in one clamping, because referencing the same supports for both operations removes a whole class of setup disagreement.
Mätning utan kontakt: Styr den optiska och datavägen
Beröringsfria metoder kan minska kontaktrisken eller möjliggöra olika förvärvsmönster, but they introduce their own conditions: target material/finish, line of sight, surface reflection, ambient light, vibration, temperatur, scan path, point density and data processing. A denser trace is not proof that the correct datum or decision rule was used.
*Engineering koncept illustration. It does not prove scanner suitability, accuracy or capability for any profile.*
Correlate to the Released Acceptance Method
Either method may be useful for process control, but neither automatically substitutes for the customer method. Compare matched representative parts using the agreed datum, support and released state. Granska repeterbarheten, bias, surface/fixture effects, calibration status and allowed difference before defining a release rule.
Use machine gauge versus customer gauge correlation, mätosäkerhet vid rakhetsinspektion och mätning av laddad kontra frigjord rakhet to plan the study.
Make the Environment Part of the Method
For non-contact stations the installation is not passive background — it is part of the measurement definition. Shrouding against ambient light, airflow and mist; isolating the optical path from press vibration; letting electronics and fixtures reach thermal equilibrium before the first reading; and re-verifying alignment after any intervention are all method steps, to be written down and scheduled like calibration. When one of those conditions changes — a new light source overhead, a relocated press, a faster line tempo — the correlation study that justified the method no longer describes the cell, and it must be repeated before its release decisions are trusted again.
Selection Trade-offs at a Glance
The table below compares the two families on the dimensions that actually decide projects. None of these rows has a universal winner — each one is a question about the specific part, tolerance and production state.
| Decision dimension | Contact route | Non-contact route |
|---|---|---|
| Surface witness risk | Present; manage with tip force, geometry and material choice | None by definition; decisive for plated, polished or coated parts |
| Sensitivity to surface finish | Low — tip integrates over texture | High — reflection and texture affect the returned signal |
| Datum explicitness | Physical supports make the datum mechanical and visible | Implied by the fixture and the software alignment; easier to leave implicit |
| Data density | Few discrete stations per trace | Dense profiles possible, but density adds processing decisions |
| Environment tolerance | Robust to light, mist and dirt typical of press cells | Ambient light, vapor, vibration and temperature drift all enter the path |
| Calibration chain | Physical references, auditable at the machine | Artifact or master-part based; verification scope must be defined |
| Correlation to customer gauge | Usually shares the same mechanical principle as acceptance gauges | Needs a defined correlation study before any release decision |
Common Selection Errors
Four errors account for most failed method choices. Första, comparing sensor datasheets instead of methods: resolution on paper says nothing about the datum, sampling plan and rule that decide the number. Andra, letting data density masquerade as accuracy — a dense scan referenced to an unrepeatable fixture produces a beautifully detailed wrong answer, and repeatability checks on reseated parts are the only honest test. Tredje, ignoring the support condition: a part measured flat on a table and accepted on centers will disagree with itself, whichever sensor is used, av de skäl som anges i loaded versus released measurement. Fourth, running the correlation study on perfect parts: matched comparison needs representative parts that exercise the tolerance band, not golden samples that everything passes.
A disciplined selection ends the way every measurement decision on this site ends: a written method — feature, utgångspunkt, stödstat, sampling, evaluation rule, acceptance limit — validated against the customer gauge on representative parts, with the study design drawn from gage R&R practice before any production claim is made.
Inputs for Method Selection
Tillhandahåll ritning/revision, required feature and limit, material/finish, dimensional range, begränsningar för delstöd, genomströmningsvillkor, miljö, strömmätare, kalibreringsbevis och kundacceptansmetod. Använd riktprovstest och acceptansguide att planera representativ bevisning, sedan kontakta StraighteningTech för ansökan diskussion.
Obligatorisk validering innan ett systemanspråk
Any measurement-system claim needs a defined method scope, kalibrerade referenser, repeated readings on representative parts, and correlation against the customer’s gauge. A generic machine capability is not a verified result.
FAQ
Is non-contact measurement always more accurate than contact?
Inga. Accuracy is a property of the whole method — datum, fixtur, sampling, algorithm and environment — not of the sensing principle. A contact system with a repeatable mechanical datum routinely outperforms a non-contact system fighting reflections and vibration. Accuracy claims require a study on the actual part, in the actual cell, against the actual acceptance rule.
When is contact the only defensible route?
When the acceptance gauge itself is mechanical, when the part must be measured in the same setup as correction, or when the production environment cannot be controlled for optics. In those cases the burden shifts to managing witness risk through tip selection and force, which is a solvable tooling problem.
Can both methods run on the same line?
Ja, and it is common: contact stations close the correction loop at the machine, while a non-contact station audits released parts downstream. The condition is a documented correlation between the two, refreshed after any change to fixtures, software or part revision, so that the two loops never drift into silent disagreement.
Relaterade StraighteningTech-resurser
Se hur automatisk axelriktning fungerar, rätningsprov test och acceptans och axelns rakhet vs runout vs TIR för generiska kontrollgränser som gäller före något kapacitetsanspråk på Kontakt kontra icke-kontakt rakhetsmätning.
*Illustration av teknikkoncept.*