Metallography · hardness · mechanics · optics — one chainMetallography & testing
National High-TechShanghai SME2 Patents

应用案例库

28 个真实工况:工件、难题、做法、依据标准、结果,五段写清。金相制样类案例占了一半——因为大部分"测不准"其实是"没制好"。

TYPICAL SCENARIOS

28 parts, 28 ways to measure

Each entry follows public standards: the part and the problem first, then how we measure, which standard applies, and what came out. Deliberately free of marketing language — compare it with your own part.

01Automotive

Carburised gears: disputed case depth

Part
20CrMnTi carburised gear, module 3, 0.6–0.9 mm required
Problem
Supplier measured 0.82 mm, OEM re-checked 0.61 mm — two verdicts on one batch, line stopped pending arbitration.
Approach
ZHVST-30F micro-Vickers system with motorised stage; one point every 0.05 mm from the surface per GB/T 9450, depth taken at 550 HV. Specimens cut on the ZQ-300F, mounted in the ZXQ-50, polished on the ZMP-3000.
Standard
GB/T 9450, ISO 2639
Outcome
Both parties re-measured on the same machine, specimen and parameters; spread narrowed to 0.03 mm and the case closed in three days without rejecting the batch.
02Bearings

Rings: sampling rate cannot keep up

Part
GCr15 bearing rings, 50,000 pcs/month
Problem
Only three parts were hardness-tested — under 5% coverage — and soft spots surfaced at assembly.
Approach
HRS-150 digital Rockwell tester gives a direct HRC reading in 8 s per part; suspect parts go to the ZQ-300F for sectioning and the 4XC-ST for structure review. Hardness distribution logged per furnace load.
Standard
GB/T 230.1, GB/T 18254
Outcome
Coverage raised to 30% while cost per part fell; soft batches are now stopped in heat treatment.
03Aerospace

Thermal-spray coatings: cross-section hardness and bonding

Part
Superalloy substrate with WC-Co / ceramic coating, 150–400 μm
Problem
Surface hardness alone cannot reveal internal porosity or interfacial cracks, so spalling stayed unexplained.
Approach
ZQ-300F low-speed precision cutting prevents edge chipping, ZXQ-50 cold mounting protects the edge, ZMP-3000 prepares the section stepwise, and the 4XC-ST inspects the interface. ZHVST-30F with 50 gf measures every 20 μm across the section.
Standard
GB/T 4340.1, ASTM E384, GB/T 9796
Outcome
A three-zone hardness profile (coating, interface, substrate) quantified the porous layer, giving spray parameters a comparable basis.
04Fasteners

High-strength bolts: decarburisation and brittle failure

Part
Grade 10.9 M12 bolts, quenched and tempered
Problem
Brittle fracture after three months in service with no visible defect — decarburisation suspected but unproven.
Approach
ZQ-300F sections along the axis, ZXQ-50 protects the thread root, ZMP-3000 polishes, and the 4XC-ST measures decarburisation depth; ZHVST-30F runs a 20 gf gradient from the surface. Judged to GB/T 3098.1.
Standard
GB/T 3098.1, ISO 898-1, GB/T 224
Outcome
Full decarburisation of 0.06 mm at the thread root confirmed an atmosphere problem; 120,000 bolts in the batch were screened.
05Mould & die

H13 cavities: hardness lost during polishing

Part
H13 die-casting cavity, 46–50 HRC required
Problem
Occasional early washout after polishing, yet every incoming hardness report passed — the culprit process stayed hidden.
Approach
HVS-5 Vickers tester takes 1 kgf points on the cavity surface and on polished areas; ZMP-3000 re-polishes and the 4XC-ST checks for tempering from overheating. Hardness compared at identical positions before and after polishing.
Standard
GB/T 4340.1, NADCA #207
Outcome
Excess polishing pressure had softened the surface by 3–4 HRC; switching to low pressure and multiple passes cut washout by 70%.
06Third-party labs

Angle metrology: traceable autocollimator readings

Part
Granite surface plates and machine guideways, straightness / flatness
Problem
Customers asked for an angle certificate with uncertainty; the existing optical instruments could not be traced.
Approach
ZY-20 autocollimator (±60″ / 0.2″) with a bridge plate in steps, straightness evaluated by minimum zone; a polygon prism cross-checks the instrument to build an internal calibration chain.
Standard
JJF 1110, GB/T 20920, JJG 117
Outcome
Certificates are accepted directly by customers, the internal calibration interval went from fixed half-yearly to on-demand, and outsourcing costs fell by two thirds.
07Forging

Large shafts: cutting distortion scraps the specimen

Part
42CrMo forged shaft, Ø180 mm × 1.2 m
Problem
An ordinary abrasive saw left blue temper colours and a 1.5 mm taper; 3 mm had to be ground away and the specimen was effectively lost.
Approach
Moved to a ZQ-500F with a large wheel, dual coolant and 1.5 mm/min feed, sectioned in steps with both ends supported; faces finished on a ZY-200.
Standard
GB/T 13298, ASTM E3
Outcome
Taper within 0.2 mm and no temper colours; preparation time fell from 40 to 15 minutes per piece.
08Battery & electronics

Electrode foils: any blade curls the edge

Part
8 µm copper and 12 µm aluminium foils, coated and uncoated zones
Problem
Abrasive cutting curled the foil and peeled the coating, so thickness and interface could not be measured.
Approach
ZY-100E diamond wire saw at low feed with no mechanical squeezing, held in a chilled fixture, then cold mounted.
Standard
GB/T 13298, ISO 19700
Outcome
Straight cut, no curl, intact coating interface; coating thickness and interface defects became measurable.
09Cemented carbide

YG grades: cobalt pools dragged out by polishing

Part
YG8 inserts, 1.5 µm grain size
Problem
Hand polishing with too much force removed cobalt and left pits that were read as porosity; batches did not agree.
Approach
A ZMP-3000 one-touch program with resin-bond diamond discs, low force, short steps and a 1 µm diamond finish, verifying each step under the microscope.
Standard
GB/T 3488, ISO 4499
Outcome
Cobalt pools preserved and porosity ratings repeatable within 0.2 of a class across three specimens.
10Heat treatment

Nitrided layers: the edge is always rounded away

Part
38CrMoAl nitrided piston rod, case 0.25 mm
Problem
After mounting the edge was chamfered and the compound layer vanished, so case depth read 30% low or not at all.
Approach
ZXQ-50 hot mounting to protect the edge, ZMP-1000 with soft cloths below 10 N for 60 s per step, ZD-300 vibratory polishing for the final finish.
Standard
GB/T 11354, ISO 6336
Outcome
Edge retention clearly improved and case depth repeated within 0.02 mm.
11Fasteners & wire

M6 bolts: the thread root disappears

Part
Grade 10.9 M6 bolts, 1 mm pitch
Problem
Specimens were too small to hold and the root was polished flat, so decarburisation and folds could not be judged.
Approach
Six pieces mounted at once in ZXQ-50 with phenolic at 140 °C and 25 MPa, ground to the pitch diameter and etched.
Standard
GB/T 3098.1, GB/T 224
Outcome
Root profile preserved, decarburisation and folds visible, 30 minutes for a batch of six.
12Casting

Heavy castings: bulk hardness will not settle

Part
HT250 machine-tool bed, 1.8 t, critical faces already machined
Problem
A portable Leeb gauge scattered on the rough face and disagreed with the foundry by 30 HBW.
Approach
HBT/HBST-3000 with a prepared Ø60 flat, 10 mm ball, 3000 kgf, 15 s dwell, indentation read by the image system.
Standard
GB/T 231.1, ISO 6506-1
Outcome
Agreement within ±10 HBW; acceptance no longer depends on feel.
13Cutting tools

Coated tools: coating and substrate hardness must be separated

Part
Carbide end mill with 3 µm TiAlN coating
Problem
Measuring at 1 kgf punched through the coating into the substrate, so the coating was never characterised.
Approach
HVF-1000D at 0.05 kgf for the coating and 0.5 kgf for the substrate, with automatic indentation measurement and photo records; coating thickness measured in the same field.
Standard
GB/T 4340.1, ISO 6507-1
Outcome
Coating 2800 HV0.05 and substrate 1600 HV0.5, both traceable and comparable between deliveries.
14Springs

Spring steel: decarburisation halves fatigue life

Part
60Si2Mn coil spring, 12 mm wire
Problem
Bench testing failed at 80,000 cycles from a surface origin while hardness passed, leaving no explanation.
Approach
Sectioned on a ZQ-50S, mounted in ZXQ-50 and traversed on an HVS-1000 every 0.05 mm, with 4XC-ST examination of the fully and partially decarburised zones.
Standard
GB/T 224, GB/T 18983
Outcome
Fully decarburised 0.05 mm and partially 0.12 mm; the atmosphere dew point was corrected and life returned to 300,000 cycles.
15Machine tools

Gantry guideways: a 3 m straightness dispute

Part
X-axis guideway, 3.2 m long
Problem
Spirit-level readings drifted with temperature and handling, and the builder and user disagreed by 0.02 mm/m.
Approach
ZY-30 autocollimator with a 200 mm bridge, 16 equally spaced points measured forward and back and evaluated by the minimum-zone method after a 2 h warm-up run.
Standard
GB/T 11336, ISO 230-1
Outcome
Both parties measured 0.018 mm full length, agreeing within 0.004 mm, and the machine was accepted first time.
16Construction

High-strength bolts: failure in the gripped thread

Part
M20 grade 10.9 bolts, tensile strength ≥1040 MPa
Problem
Specimens broke in the gripped thread, so the test was invalid even after changing grips.
Approach
WAW-600 hydraulic universal machine with bolt grips and an extensometer for Rp0.2, loaded at 10 MPa/s to GB/T 228.1.
Standard
GB/T 228.1, GB/T 3098.1
Outcome
Failure in the shank, 1085 MPa tensile and 940 MPa Rp0.2, usable for routine inspection.
17Welded structures

Low-temperature steel: impact energy falls off fast

Part
Q345E welded joint for −40 °C service
Problem
Room-temperature impact of 120 J passed, but without a transition curve the −40 °C behaviour was unknown.
Approach
JB-300B with a low-temperature bath, three specimens every 10 °C from +20 to −60 °C, plotted as energy against temperature.
Standard
GB/T 229, ISO 148-1
Outcome
Full transition curve obtained; 47 J at −40 °C against a 34 J requirement, and the welding procedure was frozen.
18Powder metallurgy

Sintered parts: porosity mistaken for inclusions

Part
Iron-base PM gear, 7.0 g/cm³ density
Problem
In bright field both pores and inclusions are black, so porosity was rated as inclusions and density was rejected.
Approach
4XC-ST bright and dark field of the same field: pores glow in dark field and change with focus, inclusions do not; porosity then quantified by image analysis.
Standard
GB/T 5163, ISO 2738
Outcome
Separated true inclusions at class 0.5 from 6.8% porosity and avoided a wrongful rejection.
19Heat exchangers

Thin-wall tubes: flattened by the vice, ovality lost

Part
20# steel heat-exchanger tube, φ19×2 mm, 300 mm long
Problem
Vice clamping plus abrasive cutting ovalised the tube end; wall thickness and roundness were unusable, while wire cutting could not get through.
Approach
ZQ-1000S auto-feed cutter with V-soft jaws and an internal mandrel, finer wheel, 0.8 mm/min feed, single-pass cutting without clamping distortion.
Standard
GB/T 13298, GB/T 246
Outcome
Ovality within 0.1 mm, wall thickness measurable, 3 min per cut.
20Automotive

Carburised gears: blue tint made the case look shallow

Part
20CrMnTi carburised and hardened gear, 0.8 mm effective case
Problem
Insufficient cooling left temper colours at the cut; the case edge was softened and measured only 0.5 mm.
Approach
ZQ-300F with dual coolant jets straight into the cut, feed reduced to 1.2 mm/min with step feeding; 0.2 mm lightly ground off before measuring.
Standard
GB/T 9450, ISO 2639
Outcome
No temper colours; case re-measured 0.78 mm, matching the heat-treatment report.
21Surface engineering

Thermal-spray coatings: the whole layer lifted off

Part
WC-12Co coating, 300 μm, on 45# steel
Problem
Hot mounting opened the coating/substrate interface and the edge broke away during grinding, so thickness and porosity could not be measured.
Approach
Vacuum impregnation plus cold epoxy mounting, 24 h cure; wire-cut sectioning; minimum pressure and a final 1 μm step of only 20 s.
Standard
GB/T 9790, ASTM B748
Outcome
Interface intact, coating 305 μm, porosity 1.2%, no interfacial cracks.
22Non-ferrous

Copper: polished surface looks fogged

Part
T2 copper plate, 3 mm, grain size 30 μm
Problem
Mechanical polishing left a foggy surface with blurred boundaries; more pressure only smeared the second phase.
Approach
ZMP-1000 low-pressure program: 1200# then 3 μm diamond for 25 s and 0.05 μm silica for 10 s; etched in ferric chloride/HCl for 8 s.
Standard
GB/T 13298, GB/T 6394
Outcome
Mirror surface, clear boundaries, grain size 8.5, repeatable across three runs.
23Aerospace

TC4 titanium: scratches that will not go away

Part
TC4 titanium forging, 12 mm thick
Problem
Titanium smears and conducts heat poorly, so papers clogged quickly and every pass added scratches; water rinsing left drying marks.
Approach
ZMP-3000 with fresh papers, low pressure and 40 s per step; 3 μm diamond on short-nap cloth, ethanol rinse and warm-air dry.
Standard
GB/T 6611, ASTM E3
Outcome
Scratches removed, α/β interfaces clear, no drying marks.
24Welding consumables

Fine wires: one at a time is too slow

Part
φ0.8 mm flux-cored welding wire, 20 sections per batch
Problem
Single mounting was slow and the wires sat at an angle, so the polished plane missed the centre.
Approach
A shop-made 12-hole fixture loaded 12 wires at once into the ZXQ-50 hot press; the fixture keeps them parallel so one grind reaches all centres.
Standard
GB/T 13298, GB/T 17493
Outcome
12 wires per press cycle, perpendicular sections, 20 wires in 40 min.
25Fasteners

Grade 12.9 bolt: decarburised depth kept measuring short

Part
M12 grade 12.9 bolt, hardened and tempered
Problem
The decarburised layer at the thread root is only tens of microns; indents spaced too widely missed the turning point and the depth was under-reported.
Approach
ZQ-50S precision cutting along the thread root, ZXQ-5A hot mounting for edge retention, ZMP-2000 to 1 μm; HVT-1000A HV0.5 indents every 0.05 mm from the root inwards, full + partial decarburisation judged per GB/T 224.
Standard
GB/T 224, GB/T 3098.1
Outcome
Full decarburisation 18 μm, total 42 μm — matching the metallographic cross-check, batch accepted.
26Advanced ceramics

Zirconia: too much polish pressure changed the hardness

Part
Y-TZP zirconia, 3Y-stabilised, 10×10×5 mm sintered tiles
Problem
Heavy polish pressure induced a tetragonal-to-monoclinic transformation layer; HV readings scattered low, giving "two hardnesses" in one batch.
Approach
ZY-100E low-speed diamond cutting, cold epoxy mounting; ZMP-2000 low-pressure long polish to 1 μm, HV1 measured unetched; indentation KIC for fracture toughness.
Standard
GB/T 16534, GB/T 6569
Outcome
Hardness stable at 1250±30 HV1, indentation KIC = 5.8 MPa·m^0.5, batch scatter under 5%.
27Electronics

MLCC capacitors: dielectric layers impossible to count

Part
0402 MLCC, ~2 μm dielectric layers, 300+ layers
Problem
Chipping at the cut face fractured every layer in that zone — no way to count layers or measure thickness.
Approach
ZQ-50S slow precision cut, vacuum epoxy infiltration 12 h, ZMP-2000 stepwise diamond polish; layer thickness distribution measured at 1000× on the 6XC-ST image analysis system.
Standard
GB/T 13298, IEC 60384
Outcome
Intact cut face, dielectric thickness 1.9–2.2 μm uniform, no interlayer cracks found.
28Gears

Carburised gear: effective case depth acceptance

Part
20CrMnTi carburised gear, module 4, case target 0.9 mm
Problem
The first sample was cut with a burnt face; the hardness gradient was wrong from the first point and the report was rejected.
Approach
ZQ-500F low-speed constant-torque cutting with flood cooling; HRC spot checks on the cut face before sectioning; HVST-1000C HV1 indents every 0.1 mm from the flank, effective depth at 550 HV1 per GB/T 9451.
Standard
GB/T 9451, GB/T 34807
Outcome
Effective case depth 0.86–0.92 mm, core hardness 42 HRC, smooth gradient, lot accepted.

These are typical conditions compiled from public standards, without customer names or commercial data. Samples of comparable parts are welcome for testing.

Compare tray 0/4
Compare now