Non-destructive Assessment of Tungsten Carbide-Backed PCD-Steel Brazed Joints Using Active Infrared Thermography
Paper details
Section titled “Paper details”| Field | Value |
|---|---|
| Published | 2026-02-26 |
| Authors | Martin Kahlmeyer, Ammar Ahsan, Christian Wolf, Igor Kryukov, Eugen Prints |
| DOI | 10.17170/kobra-2026072712307 |
Thermography separated reference and defective joints
Section titled “Thermography separated reference and defective joints”Reference joints averaged about 3.5 kN in push-off shear testing, and every defect group had more than a 50% lower failure load. Active infrared thermography separated the reference and all three defect groups through their PCD heating curves. The test used five specimens per condition and three deliberately introduced defect types. Unlike destructive push-off testing, the thermal method leaves the joint intact.
Specimens and controlled defects
Section titled “Specimens and controlled defects”The test articles were WC-8%Co-backed mixed-grain PCD saw teeth cut from a disc by wire electrical discharge machining. The diamond contained 30 µm and 2 µm grains, and each tooth was 2 mm thick. Sandblasting preceded five minutes of ultrasonic ethanol cleaning. Teeth remained in ethanol until flux application.
The filler was Ag 449 foil with a thickness of 0.1 mm. Its composition was 49% Ag, 16% Cu, 23% Zn, 7.5% Mn, and 4.5% Ni. Brazetec Special-h flux covered the tooth backs and seats. The prepared carbide and steel surfaces had mean Ra values of 2.2 µm and 1.5 µm.
Five samples were brazed for each condition below. The frequency was 20 kHz in every heating phase.
| Condition | Introduced condition | Heating schedule | Structure found after brazing |
|---|---|---|---|
| C1 | None, reference | Phase 1: 12 s at PWM 600‰; phase 2: 3 s at PWM 500‰ | Continuous filler with small isolated pores |
| C2 | Insufficient temperature | Phase 1: 9 s at PWM 600‰; phase 2: 1.5 s at PWM 500‰ | Incomplete filling and local unbonded regions |
| C3 | SAE 304 foil between tooth and filler, 0.05 mm thick and 1 mm wide | Same as C1 | Separation at the foil position |
| C4 | 0.5 mm air gap between tooth and seat | Same as C1 | Large continuous voids |
Brazing and internal defect validation
Section titled “Brazing and internal defect validation”Brazing took place in an ambient atmosphere with a 20 kHz, 10 kW induction generator and a one-sided coaxial coil. The coil sat 1 mm from the saw band. The pyrometer targeted the carbide side of the interface, where the setpoint was 720°C. The Ag 449 melting range was 680 to 705°C. A SCARA robot positioned the tips with 0.02 mm positional accuracy and 1° angular accuracy.
The authors used visual inspection, micro-computed tomography, and cross-sectional scanning electron microscopy to characterize each joint. X-ray scans used 140 kV, 10 W, and an approximately 13 µm voxel size. Each specimen supplied 1601 projections over a 360° rotation. SEM imaging used a 30 kV accelerating voltage.
C1 showed continuous filler with small isolated pores. C2 showed incomplete filling and local unbonded regions at the filler-to-carbide interface. The authors proposed insufficient alloy melting as a likely cause. C3 separated the hard metal from the filler at the inserted foil. C4 contained large continuous voids matching the air gap.
Visual inspection missed the thin C3 foil. SEM did not clearly expose C2 bonding loss. This contrast motivated a measurement based on heat transfer through the complete joint.
Active infrared acquisition
Section titled “Active infrared acquisition”Thermography used a FLIR SC5600-M with an InSb detector for mid-wavelength infrared radiation. Its spectral range was 2.5 to 5.1 µm, and its specified noise equivalent temperature difference was below 25 mK. The study recorded at 200 Hz with a 159 by 127 pixel sub-array. After internal image scaling, the effective analyzed resolution was typically about 200 by 350 pixels.
The already brazed teeth were reheated through a 20 kHz induction coil at a PWM level of 250‰. Heating lasted 2 seconds. The camera then recorded the first five seconds of cooling, giving 7 seconds total. It stood 410 mm from the tooth ends on the side opposite the coil. This transmission arrangement placed the camera and induction coil on opposite sides of the specimen.
Thermal response exposed interface resistance
Section titled “Thermal response exposed interface resistance”Steel heated faster than carbide because the steel was ferromagnetic. The authors followed heat from steel through filler and carbide into PCD. PCD had about seven times the thermal conductivity of the carbide layer. A continuous joint therefore transferred heat efficiently to the homogeneous dark PCD surface. Poor wetting, an inserted foil, or a gap added thermal resistance and delayed the PCD response.
The measured quantity was DI, the temporal change in infrared intensity at a pixel or region. Since infrared intensity tracks surface temperature, a delayed DI response indicated reduced heat transfer through the interface. Defects produced little localized contrast in individual frames. The defects were small relative to the heated area, and the induction coil created a dominant global thermal field. The authors therefore relied on time-dependent curves rather than single thermograms.
A line object along the PCD layer supplied an average intensity at each time step. On the double-logarithmic presentation, the C1 reference heated fastest. The poor C2 bond reduced the heating rate. The C3 foil also reduced the heating rate because its conductivity was lower than that of the silver-based filler. The C4 air gap caused the largest delay and was distinct from the other defect patterns. The authors report that all three defective configurations could be differentiated from their thermal response.
The authors also propose local intensity thresholds against a reference joint. Such thresholds could flag an irregularity and help infer its type. The paper reports no numerical thresholds or validation of that threshold method.
Push-off testing confirmed the strength loss
Section titled “Push-off testing confirmed the strength loss”Push-off testing pushed each tooth from the carbide side while a clamp held the steel carrier. Mechanical testing made all three defect types clear, although the individual non-destructive imaging methods did not.
C2, C3, and C4 averaged around 1.6, 1.8, and 1.0 kN. The air gap produced the lowest mean load. The foil insert retained the highest mean among defective joints. These results support the thermal response as an indicator of joint quality. The paper does not report a correlation coefficient between thermographic curves and shear load.
Limits for inline quality assurance
Section titled “Limits for inline quality assurance”The results support feasibility rather than complete production qualification. The sample set was limited to five deliberately defective or reference samples per condition. The authors did not report acceptance limits, classification accuracy, false calls, inspection speed, or production variation beyond the tested bandsaw geometry.
Thermography followed brazing in this experiment and used reheating. The authors propose transferring the evaluation to inspection during induction brazing because the same heating principle applies. The proposed in-process use would require validation under changing production conditions.
The paper cites acoustic emission, four-point electrical resistance, ultrasonic testing, pulsed eddy current, micro-computed tomography, thermal probes, and modal analysis as earlier quality-assessment approaches. It identifies no rapid, accurate, robust method for high-throughput inspection of WC-backed PCD-to-steel joints. The authors state that this is the first thermography application for defect detection in induction-brazed cutting elements.
For saw-blade production, the temporal curve provides a candidate screening signal for joint quality. The authors also propose using it to monitor brazing parameters. No production acceptance rule or industrial validation is established here. Further studies are required to define reliable evaluation methods for production use.
View Original Abstract
Brazing is the most commonly used method for joining carbide and carbide-coated polycrystalline diamond inserts to steel bodies of metal and wood saw blades. However, there is currently no non-destructive inline method for evaluating joint quality. This may lead to insufficient tool performance and high costs, particularly for large saw blades with many teeth. To address this challenge, we present a novel in-line testing method designed to reliably and efficiently assess brazing quality. Polycrystalline diamond inserts with carbide backing were inductively brazed to steel surfaces using a silver-based brazing alloy. Artificial defects, including incorrect positioning, inclusions, or deliberately chosen incorrect brazing parameters, were introduced into some of the brazed joints. The heat transfer rate during brazing was monitored through thermography. The proposed method reliably identified artificially introduced defects in brazed joints and showed good correlation with destructive push-off shear testing. These findings demonstrate the potential of induction-excited thermography for inline quality assurance of brazed joints in industrial saw blade production.
Advanced Thermal Simulation Service Diamond Heat Spreader Performance Analysis
View Specs →
Diamond Heat Spreader With Copper Coated Metalized Surface
View Specs →