よくある質問

よくある質問

選定ガイド · 試験方法

High-temperature Hardness Testers & Indenters

Why must hot hardness be measured in vacuum or inert atmosphere?

Above 600°C the surface oxidises quickly: the oxide layer is far softer than the substrate and thickens with time, so indentation diagonals no longer reflect the base material. Diamond indenters also begin to graphitise above 700°C in air.

The chamber is therefore pumped to ~10⁻² Pa or purged with high-purity argon keeping oxygen below 50 ppm.

Can hot hardness be converted to room-temperature hardness?

No. At temperature the material creeps and the indentation keeps growing with dwell time, so hardness is a function of both temperature and time. Always report temperature, dwell time, atmosphere and indenter type.

How much error does an uneven temperature field cause?

Hardness is strongly temperature dependent — about 2–3 HV per 10°C for aluminium near 300°C. Keep the gradient within ±3°C and ≤5°C across the effective zone around the indentation.

What else can hot indentation measure besides hardness?

One indent also yields elastic modulus from the unloading slope, creep exponent from the dwell segment, indentation fracture toughness, and hardness gradients across coatings or case depth.

Which heating elements reach 1600°C, and how should samples be prepared?

SiC elements reach 1400°C, MoSi₂ reaches 1700°C, graphite works to 2000°C in vacuum or inert gas. Samples: φ20×10 mm or 15×15×8 mm, faces parallel within 0.02 mm, Ra ≤0.2 µm, free of cutting burn.

Photoelectric Autocollimators

What can a photoelectric autocollimator measure?

Micro-angles down to 0.01″. With accessories it covers straightness, flatness, parallelism, squareness and machine-tool geometry. Unlike interferometers it measures angle rather than displacement.

Resolution or accuracy — which matters more?

Resolution is only the display step. What matters is the indication error over the full range and repeatability. Check full-range error, local accuracy within ±10″, and 10-run repeatability.

Does accuracy drop with longer working distance?

Yes. Return intensity falls with the square of distance and the spot grows, raising centroid noise; air turbulence adds jitter. Use longer focal length and larger mirrors for long range.

Why is a precision flat mirror necessary?

The instrument measures the mirror normal, so mirror figure error adds directly to the result. Use λ/10 or better flats with fine pitch/yaw adjustment.

How is guideway straightness measured with an autocollimator?

Mount a mirror on the moving carriage, read angle at equal intervals, integrate angle against travel and evaluate the curve by the minimum-zone method. Measure both directions to cancel backlash.

Microscopes

Metallurgical, stereo or toolmaker microscope?

Microstructure and case depth: metallurgical. Fractures and assembly: stereo. Geometry such as holes and threads: toolmaker or vision system.

When to use brightfield, darkfield, polarised light or DIC?

Brightfield for routine structure; darkfield for cracks and second-phase edges; polarised light for anisotropic phases; DIC for low-relief steps such as case-depth transitions.

How do numerical aperture and resolution relate?

δ = 0.61λ/NA. Higher NA gives better resolution but shallower depth of field and shorter working distance. Match total magnification to about 500–1000× NA.

Why do 100× objectives need immersion oil?

Air limits NA to about 0.95; oil with n=1.515 allows NA 1.4 and roughly 1.5× better resolution. Clean the lens immediately after use.

Is a higher-resolution camera always better?

No. Pixels should match the objective resolution (Nyquist). Over-sampling only magnifies noise and lowers frame rate.

Vision Measuring Systems

What affects vision-measuring accuracy?

Scale resolution and error, lens distortion, Z-axis focusing repeatability, temperature (20±2°C) and fixturing. Look at the (3+L/200) µm style accuracy statement, not just resolution.

How does a 2.5D vision system differ from a CMM?

Vision measures 2D size and profile optically (Z by focus); a CMM uses contact or scanning probes for true 3D geometry and GD&T. Choose by part type.

When is autofocus or a laser probe needed?

When you need Z data — steps, blind-hole depth. Autofocus suits mirror-like parts; laser triangulation suits rough surfaces and fast scanning.

Repeatability vs indication error?

Repeatability is the spread of repeated measurements (stability); indication error is deviation from the true value (accuracy). Production cares about repeatability, acceptance about error.

How to choose illumination?

Backlight for outlines and through-holes; ring light for general surfaces; coaxial for mirrors and deep holes; low-angle for scratches and texture.

Testing Machines

Electronic or hydraulic universal testing machine?

Electronic (servo + ball screw) gives precise speed control up to ~300 kN; hydraulic suits high loads to 2000 kN. Under 100 kN, prefer electronic.

How to choose an extensometer?

Clip-on for high accuracy at low cost; video extensometer for high temperature, thin sheets or fragile specimens; automatic for batch testing. Yield strength requires an extensometer.

Does testing speed affect results?

Yes. Most metals are strain-rate sensitive: faster testing raises yield and tensile strength and lowers elongation. Follow ISO 6892-1 method A or B consistently.

Why do grips and alignment matter?

Eccentric loading adds bending, lowering strength and scattering results. Keep alignment within 10% and verify with an alignment fixture periodically.

How often should the machine be calibrated?

Force annually with a 0.05-class proving instrument at 3–5 points; displacement and speed annually; extensometer every two years. Add interim checks with reference specimens.

Hardness Testers

Rockwell, Brinell or Vickers?

Rockwell for fast production checks, Brinell for castings and coarse grains, Vickers for thin parts, case depth and ceramics. Use low-load Vickers for thin or hardened layers.

What thickness does the specimen need?

At least 10× the indentation depth for Vickers and Rockwell; Brinell needs ≥3 mm and 1/3 of the indentation diameter. Thin samples show an anvil effect with high readings.

Can hardness conversion tables be used directly?

Not for acceptance. Conversion tables are statistical results for specific material families. If the drawing says HRC, measure HRC.

How to test curved, thin or small parts?

Use V-blocks with curvature correction for cylinders, low-load Vickers or UCI for thin walls, and mounting/fixtures for small parts.

What daily checks are needed?

Indent a reference block 3–5 times on the working scale. Deviation beyond ±1 HRC (±3% HV) means checking indenter, force, support and temperature.

Profilometers & Roundness Testers

Difference between Ra, Rz and Rq?

Ra is the arithmetic mean, Rz the maximum height, Rq the RMS. Ra hides isolated deep valleys — add Rz for sealing, coating or fatigue surfaces.

How to choose cut-off and sampling length?

Per ISO 4288: 0.08 mm for Ra 0.006–0.02 µm, 0.25 mm for 0.02–0.1, 0.8 mm for 0.1–2, 2.5 mm for 2–10, 8 mm above. A wrong cut-off can double the result.

Profile meter vs roughness tester?

Roughness testers evaluate micro-irregularities; profile meters measure macro geometry such as radii, angles and groove depth. Combined systems give both in one trace.

What determines roundness instrument accuracy?

Spindle radial error (typically 0.02–0.05 µm) and sensor resolution, plus centring/levelling, filtering and evaluation method. Centre the part within 1/10 of tolerance.

Does measurement direction matter?

Yes — surfaces are anisotropic. Measure across the lay for turned surfaces, keep away from edges and keyways by at least 5× cut-off, and report 3 locations or more.

Coordinate Measuring Machines

How to read CMM accuracy specs?

MPEE = (1.8 + L/300) µm grows with length; MPEP is the probing error. Both matter — a poor MPEP ruins small-feature measurement.

Why does a CMM need a temperature-controlled room?

Steel expands ~1.2 µm per 100 mm per °C — the same order as machine accuracy. Specs assume 20±2°C and gradients within 1°C/h.

Touch trigger, scanning or optical probes?

Trigger probes for points and holes; scanning for free-form surfaces; optical for soft or thin parts. Multisensor probes finish everything in one setup.

How to keep measurement consistent?

Fix the alignment convention, point count and distribution, and evaluation algorithm, and document them. Differences of tens of microns between programs are common.

What routine maintenance is required?

Daily: clean guideways and probe, check air pressure, home the axes. Weekly: check scale heads and air bearing gaps. Yearly: calibrate and replace filters.

Spectrometers & Photometers

OES vs handheld XRF?

OES measures C, S, P, N with high precision but needs argon and prepared samples; XRF is portable and non-destructive but blind to light elements. Use both for different jobs.

How should samples be prepared?

Representative sampling, then grind flat with 60–120 grit alumina, avoiding scratches and overheating. Soft metals such as Al and Mg should be milled, not ground.

How often to standardise?

Type standardisation every shift or 2 hours; full standardisation weekly or after service. Always re-standardise after changing argon or cleaning the lens.

What happens with impure argon?

Oxygen, nitrogen and moisture blacken the burn spot and depress light-element results. Use 5N purity, dew point ≤-60°C and stainless or copper lines.

What detection limits are achievable?

Typically ppm level for common elements, sub-ppm with high-resolution systems. Ask for 3σ measured values on your matrix.

Metallographic Preparation

How to avoid cutting burn?

Choose the right blade, ensure copious coolant aimed at the cut, and reduce feed rate. Burn appears as a white layer that polishing cannot remove.

Hot or cold mounting?

Hot mounting is fast with good edge retention; cold mounting suits heat-sensitive or coated samples. Use vacuum impregnation for porous or coated interfaces.

What grinding/polishing sequence?

180 → 400 → 800 → 1200 grit, each perpendicular to the previous, then 3 µm and 1 µm diamond followed by 0.05 µm silica. Reduce pressure on soft metals.

How to choose an etchant?

Nital for carbon and low-alloy steels, aqua regia for stainless, Keller for aluminium, ferric chloride for copper alloys. Start dilute and short, stop at clear grain boundaries.

How to troubleshoot preparation defects?

Scratches mean incomplete previous step or dirty cloth; comet tails need lower pressure; relief needs shorter high-pH polishing; rounded edges need better mounting.

⇄0比較仕様比較