Skip to content

Investigation of the Effect of B4C Addition on the Matrix Composition of Diamond Sockets in Granite Cutting and the Relationships Between Socket Metallographic Properties and Cuttability Parameters via SEM and Regression Analysis

FieldValue
JournalMaterials
Published2026-09-24
AuthorsGökhan EKİNCİOĞLU, R. Altındağ
DOI10.3390/ma19194079

The 4% B4C socket produced the lowest specific wear among eight compositions, at 84.84 µc/m² after 60 granite cuts. It also recorded the highest mean Knoop hardness, 303.46 HK, and the highest net power consumption, 720.85 W. Its specific cutting energy was also the highest at 1.4826 J/mm³. The result creates a trade-off. The 4% matrix resisted removal but demanded more cutting energy. The authors associated this behavior with a regular diamond and matrix interface and a visually uniform dispersion of 3 to 8 µm B4C grains. They did not call this composition an unqualified optimum.

The matrix program covered eight powder metallurgy compositions with B4C from 0% to 7% in 1% increments. B4C replaced cobalt on a one-to-one weight basis. Bronze stayed at 29 wt.%, nickel at 1 wt.%, and diamond content stayed constant. Cobalt fell from 70 wt.% to 63 wt.% as B4C rose from 0% to 7%.

The sockets were sintered at 730 °C and 35 MPa for 3 min, then welded to 350 mm diameter steel discs. Unit volume weight was measured with the ASTM B962 water immersion method. Relative density and porosity used the measured value and theoretical density calculated from the powder proportions.

Each composition had three independent sockets for Knoop hardness and porosity. Fifty surface readings per socket produced 1,200 hardness measurements across the program. Abrasion loss by weight, or ALW, came from two independent sockets per composition.

SEM examined fracture surfaces with a LEO 1430 VP microscope. The authors used visual observations of pores, B4C distribution, and diamond and matrix interfaces.

Cutting trials used Hisar Gray, a quartz monzonite with 206.63 MPa uniaxial compressive strength, 445.82 HK Knoop hardness, and 67.82% SiO2. Each specimen measured 5 cm × 10 cm × 25 cm.

The authors noted that the rock’s 445.82 HK exceeded the highest socket mean of 303.46 HK. Peripheral speed stayed at 30 m/s, depth of cut at 10 mm, and feed rate at 0.7 m/min.

One saw per composition completed 60 consecutive cuts. Net power was full-load power minus unloaded power. Specific cutting energy was energy per unit volume of removed rock.

Reported net power and cutting energy values were means over those cuts. Their standard deviations describe within-test variation, not blade-to-blade variation. Specific wear came from one socket height-loss measurement per composition, so no standard deviation was calculated.

Table 1 lists the reported mean values for unit volume weight, porosity, Knoop hardness, and ALW.

B4C (wt.%)Unit volume weight (g/cm³)Porosity (%)Knoop hardness (HK)ALW (g)
08.053.48244.470.00374
17.567.24241.580.00500
27.535.52253.350.00445
37.345.90279.130.00450
47.304.20303.460.00297
56.828.70244.670.00643
66.4211.932340.00628
76.0514.312420.00768

At 4% B4C, porosity was 4.20% with SD 1.88%, and hardness was 303.46 HK with SD 65.94 HK. ALW was 0.00297 g with SD 0.0019 g.

At 0% B4C, SEM showed pores of about 2 to 4 µm and an approximately 8 µm interface gap. At 1%, pores were about 10 to 12 µm, and the interface looked better. At 3%, the matrix contained cracks of about 40 to 50 µm. At 4%, the visual distribution was most uniform, with grains measuring 3 to 8 µm and a regular interface. The matrix also fractured in a lamellar pattern. At 5%, the matrix became coarse and heterogeneous. At 6%, pores were about 35 to 40 µm. At 7%, pores exceeded 40 µm.

All SEM observations were qualitative. No quantitative image analysis, pore statistics, or EDS verification was reported.

The authors attributed higher porosity to reduced cobalt-derived liquid phase during sintering. They state that partially melted cobalt fills interparticle voids. Lower cobalt therefore reduced sinterability as B4C rose. Unit volume weight also fell because B4C has a density of 2.52 g/cm³. The other matrix components had reported densities of 8.60 to 8.90 g/cm³.

Table 2 gives the mean cutting results for each composition.

B4C (wt.%)Net power (W)Specific cutting energy (J/mm³)Specific wear (µc/m²)
0707.851.455390.34
1529.321.0493100.15
2561.861.098698.98
3570.811.263293.73
4720.851.482684.84
5636.281.1896109.70
6452.220.9144115.41
7430.260.9044128.52

The 4% saw recorded 720.85 ± 48.25 W and 1.4826 ± 0.076 J/mm³. At 7%, specific cutting energy was 0.9044 J/mm³ and net power was 430.26 W. Specific wear reached 128.52 µc/m², while porosity reached 14.31%.

The authors suggest that accelerated matrix removal exposed fresh diamond grains. They interpret this mechanism as lower energy demand with higher wear. The undoped saw also had high specific cutting energy of 1.4553 J/mm³ and net power of 707.85 W. The authors linked these values to its approximately 8 µm interface gap and poor wetting. The measured wear data show why cutting energy alone did not identify the best socket.

Separate simple regressions used the eight composition means. Porosity was the strongest common predictor across the three cutting outcomes. It had r = -0.87 with net power, r = -0.88 with specific cutting energy, and r = 0.97 with specific wear. The corresponding p-values were 0.005, 0.004, and below 0.001.

ALW had r = -0.78 with net power, r = -0.84 with specific cutting energy, and r = 0.98 with specific wear. The p-values were 0.02, 0.01, and below 0.001. Its direct relation with wear supports using ALW as a pre-cut indicator within this socket system.

Unit volume weight had positive r = 0.70 correlations with net power and specific cutting energy. Its relation with specific wear was negative, with r = -0.87.

Hardness correlations were weaker in the full set. Net power, cutting energy, and wear had r = 0.56, 0.67, and -0.69. Excluding the undoped composition strengthened them to 0.76, 0.92, and -0.83. This sensitivity analysis used seven compositions. These pairwise models do not separate the independent effects of correlated properties.

Earlier work on the same socket set measured cutting performance but did not resolve these microstructural links. A related study examined ultrasonic velocity. This manuscript uses data from the first author’s 2017 doctoral thesis and adds the regression and SEM interpretation.

The matrix design cannot isolate B4C from cobalt. Each increase in B4C caused a matching cobalt decrease, so the measured changes reflect their combined effect. The cutting evidence is also narrow. One saw per composition cut Hisar Gray at one fixed operating setting. The authors require validation across different rock types and cutting conditions before broader generalization.

Two internal reporting issues require attention. The methods specify 730 °C, while the mechanism discussion refers to 750 °C during cobalt melting. The text also calls the 4% socket the densest, although its unit volume weight of 7.30 g/cm³ is below the 1% to 3% values of 7.56, 7.53, and 7.34 g/cm³. These discrepancies should be resolved before applying the proposed mechanism to new process settings.

Porosity and ALW were the stronger practical wear indicators within this test. The 4% B4C composition offered the lowest observed wear if its higher energy demand was acceptable. It should not be treated as a universal optimum. A broader socket design should confirm hardness, porosity, and ALW in the intended rock and cutting conditions.

View Original Abstract

Selecting diamond sockets with matrix properties compatible with the target rock is critical for cutting efficiency and cost reduction in natural stone processing. This study investigated the effect of B4C addition (0-7 wt.%) on the matrix microstructure, diamond-matrix interfacial bonding quality, and cutting performance of powder metallurgy diamond sockets used in granite sawing. Eight sockets, one undoped and seven with varying B4C contents, were sintered at 730 °C, 35 MPa, for 3 min. Unit volume weight (UVW), porosity (P), Knoop hardness (HK), and abrasion loss by weight (ALW) were determined for each composition. SEM analyses on fracture surfaces revealed significant microstructural differences with increasing B4C content. Pore sizes increased from 2-4 µm (0% B4C) to over 40 µm (7% B4C). The 4% B4C socket achieved the most uniform B4C grain distribution (3-8 µm), the highest Knoop hardness (303.46 HK), the lowest porosity (4.20%) among doped sockets, and the tightest diamond-matrix bonding. Cutting tests on Hisar Grey granite confirmed that the 4% B4C socket delivered the highest net power consumption (720.85 W) and the lowest specific wear (84.84 µc/m2), consistent with SEM observations. Regression analyses identified porosity as the single strongest predictor across all cutting parameters (r = 0.87-0.97), and abrasion loss by weight as a reliable indicator of specific wear (r = 0.98). These findings suggest that laboratory metallographic measurements may serve as useful indicators of cutting performance trends, though validation across different rock types and cutting conditions is needed before broader generalisations can be drawn.