Troubleshooting Common Tungsten Carbide Manufacturing Problems

This master table covers every major defect type across the entire manufacturing process. The "Severity" column prioritizes your response: Critical defects (eta phase) mean the entire batch may be scrap; High severity defects (capping, lamination) require immediate process stops; Medium severity may allow continued production with monitoring. The "First Check" column tells you what to measure immediately—don't start changing parameters until you've collected this data. The "Quick Fix" gives you the most likely solution, but always verify the root cause first. Notice that most pressing defects relate to speed and pressure, while sintering defects often trace back to temperature or carbon balance.

This diagnostic flowchart systematically narrows down defect types based on visual observation. Start at the top: when did you find the defect? Pressing defects (found immediately after ejection) follow the left branch—crack location and type identify the specific issue. Sintering defects follow the center branch—first check for cracks, then surface appearance if no cracks are visible. Property failures found at inspection follow the right branch—the out-of-spec property points to specific root causes. Each endpoint (capping, lamination, dewax crack, etc.) corresponds to a specific set of causes and corrective actions detailed in the sections below. The crack pattern is particularly diagnostic: horizontal cracks suggest pressure or dewaxing issues, radial cracks suggest thermal shock, random cracks suggest carbon imbalance.

This quick-reference pairs with the flowchart above. Once you've identified the defect type using the flowchart, this table gives you the immediate cause and fix. Note the distinction between pressing cracks and sintering cracks: pressing cracks (capping, lamination, spring-back) happen before the part enters the furnace, while sintering cracks (dewax, thermal, carbon) develop during the thermal cycle. This distinction is critical because it tells you which process stage to investigate. Pressing cracks mean you need to examine pressing parameters, die condition, and powder properties; sintering cracks mean you need to examine the thermal profile, atmosphere, and carbon balance.

Capping is one of the most common pressing defects and is particularly frustrating because it often appears only during or after ejection—the part looks perfect in the die. The crack occurs because air becomes trapped in the compact during pressing; as the upper punch retracts and pressure releases, the trapped air expands, literally blowing off the top layer. Tall parts with high length-to-diameter ratios are most susceptible because air must travel further to escape. Fine powders with poor flowability are more prone to air entrapment. Capping is 100% scrap—you cannot salvage a capped part because the crack propagates during sintering even if initially hairline-thin.

This root cause analysis is ordered by likelihood—start your investigation at the top. Punch speed being too fast is the most common cause; the final 5mm of travel should slow dramatically to allow air to escape through the powder ahead of the advancing punch. Insufficient dwell at peak pressure doesn't allow time for air to squeeze out through remaining porosity. Die wall friction prevents the compact from sliding smoothly during ejection, creating stress concentrations. The diagnostic column tells you what to check before making changes. For punch speed, review the press program's speed profile; you should see a distinct slowdown for the final stroke. For dwell time, check cycle time records—if dwell was reduced to speed up production, that's likely your cause.

Lamination differs from capping in location and mechanism. While capping occurs at the top surface due to air expansion, lamination occurs through the body of the part due to density gradients created during compaction. The horizontal layers form at planes where density changes abruptly—typically where the pressure wave from each punch transitions. These density gradients create differential shrinkage during sintering, pulling the part apart at the weak interfaces. Lamination may not be visible after pressing; hairline cracks that you can't see open up during sintering as different density zones shrink at different rates. Parts with high L/D ratios are most susceptible because pressure transmission from punch to remote areas is inherently limited.

Over-pressure is the leading cause of lamination. When pressure exceeds the powder's capacity to transmit load uniformly, shear planes develop. Die wear creates a "shelf" where the worn section meets unworn section, interrupting smooth pressure transmission and creating a lamination plane at that exact location. Powder overfill means more material must be compressed into the same space, requiring higher pressure and increasing lamination risk. Non-uniform fill creates density differences before pressing even begins. Moisture-affected powder flows poorly and compacts unevenly. When troubleshooting, measure actual peak force and compare to successful prior runs—if force crept upward (perhaps due to die wear requiring more force to eject parts), that's your likely cause.

Vertical cracks result from spring-back—the elastic recovery of the compact as pressure is released and the part exits the die. All pressed parts experience some elastic expansion; the problem occurs when this expansion is constrained or uneven. Corners and edges are most vulnerable because stress concentrates at these geometric discontinuities. Complex shapes with varying wall thicknesses experience differential spring-back that creates internal stresses. Vertical cracks can be subtle initially but always propagate during sintering because the stress concentration remains. Unlike horizontal cracks that relate to compaction uniformity, vertical cracks relate to the relationship between elastic recovery and die constraint during ejection.

Spring-back during ejection is the primary mechanism. As the part exits the die, it wants to expand elastically, but the die walls constrain this expansion. If the part can't expand gradually as it exits (no draft angle) or if friction holds it in place until sudden release, the result is a tensile crack on the surface. Die wall friction exacerbates this by creating stick-slip behavior—the part doesn't slide smoothly but rather sticks then jumps, creating peak stresses. Adding draft angle (0.25-0.5° taper with larger dimension at top) allows gradual expansion as the part ejects. Polishing the die reduces friction. If you measure a part immediately after pressing and again after 24 hours, you can quantify elastic recovery—if it exceeds 0.5%, vertical cracking risk is high.

Low green strength means the pressed part lacks sufficient cohesion to survive handling before sintering. Unlike the crack defects above which create discrete fractures, low green strength manifests as general fragility—parts crumble at edges, shed powder when touched, or break during transfer to sintering trays. The binder system (typically paraffin or PEG) provides green strength by acting as glue between particles. If binder content is low, binder is degraded from improper storage, or pressing conditions don't activate the binder properly, green strength suffers. Coarse powders inherently have lower green strength than fine powders because fewer particle-to-particle contacts exist per unit volume. This defect is rated "medium" severity because careful handling can sometimes get parts to the furnace intact.

Low pressure is the most common cause—if green density is below target, there aren't enough particle contacts to provide mechanical interlocking, and binder alone can't hold the structure together. Binder degradation from improper storage (high humidity, temperature cycling, extended storage after opening) is equally common and often overlooked. Check the powder's age and storage history before blaming the pressing process. The COA should list binder content; if actual binder is below specification, contact the supplier. Coarse powders have inherently lower green strength and may need higher pressing pressures than fine powders to achieve adequate handling strength. Moisture contamination interferes with binder function and should be checked via Hall flow rate—degraded powder flows slowly or not at all.

The ASTM B276 porosity classification system divides pores into three types based on size and morphology, which correlates with different root causes. Type A porosity is small (<10 μm), uniformly dispersed, and results from incomplete densification—either temperature was too low, hold time was too short, or green density was insufficient to achieve full density. Type B porosity is larger (10-25 μm), irregular in shape, and typically indicates carbon imbalance causing gas evolution or contamination introducing foreign phases. Type C porosity is very large (>25 μm), often angular, and indicates free carbon or graphite contamination—the graphite burns out during sintering leaving large voids. This classification is critical for troubleshooting because each type points to a different process issue.

The ASTM B276 rating system quantifies porosity severity for specification and quality control. The rating is determined by comparing a polished metallographic section to standard photographs—each grade represents a specific pore population. A00 is essentially perfect; A02 allows a few tiny pores; A04 is the typical specification for standard carbide grades. A06 and A08 indicate progressively more porosity and typically result in reduced mechanical properties—A08 is often a reject condition. The B and C ratings indicate specific problem types (carbon-related and free carbon) regardless of size, because these indicate process problems that should be corrected. When reviewing metallography, note both the letter (porosity type) and the number (severity); "A04 B02" would indicate slight normal porosity plus trace carbon-related porosity.

This root cause matrix guides you from porosity type to corrective action. Type A porosity (under-sintering) has the most straightforward fixes: increase temperature, extend hold time, or improve green density to reduce the densification distance. Type B porosity requires carbon balance investigation using magnetic saturation as the primary diagnostic—values below 80% indicate carbon deficiency (eta phase forming, releasing gases), while values above 95% suggest carbon excess (free carbon creating voids). Type C porosity with its large angular voids usually indicates graphite contamination from furnace components, setters, or handling equipment—inspect and clean all potential sources. Always address the root cause rather than just compensating with higher sintering temperature, which risks grain growth.

Eta phase is the most feared defect in WC-Co production because it's irreversible and dramatically reduces mechanical properties. When carbon is deficient, tungsten and cobalt react to form complex carbides: M₆C (Co₃W₃C) or M₁₂C (Co₆W₆C). These phases are extremely hard but brittle, acting as stress concentrators that initiate cracks under load. Even 1-2 volume percent eta phase can reduce transverse rupture strength by 30-50%. The angular appearance in polished sections is distinctive—unlike rounded porosity, eta phase has sharp geometric facets. Magnetic saturation below 80% of theoretical value is the primary screening test because cobalt bound in eta phase loses its ferromagnetic properties. Once formed, eta phase cannot be removed by heat treatment; the batch is scrap.

Carbon deficiency has two sources: the powder started with insufficient carbon, or carbon was lost during sintering. Powder carbon content should be verified against the COA and compared to previous successful lots. Decarburization during sintering occurs when oxygen in the atmosphere or on the powder surface reacts with carbon to form CO gas, which is pumped away. Poor vacuum, high oxygen content in the powder, or ceramic (rather than graphite) furnace furniture all promote decarburization. Over-sintering at high temperature or long time increases decarburization exposure. The corrective actions are straightforward: either use powder with higher carbon content or reduce decarburization by improving vacuum and using graphite furniture. Always run magnetic saturation on sintered samples before shipping.

Grain growth occurs when WC particles coarsen during sintering through the solution-reprecipitation mechanism: small grains dissolve into the liquid cobalt and redeposit on larger grains. This is thermodynamically favorable because larger grains have lower surface energy. The result is a coarser microstructure than the starting powder, reducing hardness according to the Hall-Petch relationship. Fine-grain powders (submicron) are most susceptible because grain growth is exponential with temperature and time—the same sintering cycle that works for coarse powder may cause excessive coarsening in fine powder. Grain growth is "medium" severity because the parts may still function, just with reduced hardness; however, for precision applications where hardness is specified, grown parts must be rejected.

Temperature and time are the primary drivers of grain growth because dissolution and reprecipitation rates increase exponentially with temperature. Even 20°C above optimal can double grain growth rate in fine-grain grades. Hold time matters because grain growth continues as long as liquid cobalt is present—longer holds mean more growth. Grain growth inhibitors (vanadium carbide VC, chromium carbide Cr₃C₂) are added to fine-grain powders specifically to retard grain growth; if a non-inhibited powder was used accidentally, severe grain growth will occur at normal sintering conditions. When investigating grain growth, compare the actual sintering cycle to the qualified cycle for that specific powder grade—each grade has an optimal temperature range.

Coercivity (magnetic coercive force) is the fastest way to detect grain growth without destructive metallographic examination. Coercivity decreases as grain size increases because larger grains have fewer domain wall pinning sites. Each powder grade should have a coercivity specification based on the expected sintered grain size. A 5-10% reduction from target indicates slight grain growth that may be acceptable depending on application requirements. 10-20% reduction indicates moderate growth that probably affects hardness. Over 20% reduction indicates severe growth—sintering temperature or time was significantly out of control. Establishing coercivity limits for each grade and checking every batch provides continuous process monitoring.

Cobalt sweating occurs when liquid cobalt, driven by capillary forces and gravity, migrates to the part surface during sintering. The shiny metallic beads on the surface are pure cobalt that has left the interior, creating cobalt-depleted zones that are harder but more brittle than specification. The surface cobalt is soft and will wear rapidly in service. Worse, the internal cobalt depletion creates a gradient in properties—the part may meet hardness specs at the surface but have entirely different properties in the core. This defect is particularly problematic for parts with functional surfaces because the service performance will differ dramatically from qualification testing.

Peak temperature is the primary cause because cobalt mobility increases dramatically above the optimal sintering range. Just 20-30°C above specification can cause visible sweating. High carbon content lowers the eutectic temperature (the point where liquid cobalt forms), effectively making the sintering temperature "higher" relative to the eutectic. Excessive cobalt content provides more liquid phase that is more prone to migration. Poor vacuum can contribute by allowing oxide films that affect wetting behavior. When troubleshooting, first verify actual peak temperature with a thermocouple survey—furnace controllers can drift, and parts at different locations experience different temperatures. A 20-40°C reduction usually resolves the issue.

Distortion occurs when different regions of a part shrink by different amounts during sintering. Since WC-Co shrinks approximately 17-20% linearly during sintering, even small percentage differences in shrinkage between regions create significant shape changes. A 1% difference in local shrinkage across a 50mm part creates nearly 0.5mm distortion. Thin parts and long slender parts are most susceptible because they have low stiffness to resist distortion forces. Unlike some defects that become obvious immediately, distortion may only become apparent during final dimensional inspection, making it particularly frustrating—you've invested full processing cost before discovering the problem.

Green density gradient is the leading cause: if one side of the part is pressed to higher density than the other, that side will shrink less during sintering, causing bowing toward the denser side. This can result from non-uniform die fill, asymmetric punch action, or die wear creating friction differences. Uneven heating causes similar effects—hotter regions sinter faster and can "set" the structure while cooler regions are still densifying, creating internal stress. Improper support allows gravity to pull on softened parts during the liquid phase sintering stage. Long parts must be fully supported along their length to prevent sagging. Sectioning and measuring green part density at multiple locations reveals whether the problem originates in pressing or sintering.

Hardness out of specification requires understanding what drives hardness in WC-Co: WC grain size (finer = harder) and cobalt content (less Co = harder). Low hardness can result from grain growth (grains larger than powder specification), high cobalt (more soft binder phase), or eta phase (which appears hard locally but creates soft cobalt-depleted regions nearby). High hardness typically indicates low cobalt or porosity—porosity can artificially elevate hardness readings because the indenter encounters less binder. Variable hardness across a part indicates inhomogeneous microstructure from poor mixing or uneven sintering. The diagnostic tests in the middle column identify which mechanism is responsible.

Transverse rupture strength (TRS) is the most sensitive indicator of structural defects because it measures resistance to crack propagation. Each defect type reduces TRS by a characteristic amount. Type A porosity (under-sintering) causes 10-20% reduction—the pores act as stress concentrators. Type B porosity (carbon-related) is more damaging at 20-40% reduction because the pores often have angular shapes. Eta phase is devastating at 30-50% reduction because these hard, brittle phases initiate cracks under tensile stress. Grain growth reduces TRS 10-30% because larger grains have longer slip planes. Inclusions and surface defects create stress concentrations that nucleate fracture. When TRS is low, work through this list systematically using the diagnostic tests.

Magnetic properties are the primary quality control tools for WC-Co because they're fast, non-destructive, and correlate directly with microstructure. Magnetic saturation measures the amount of ferromagnetic cobalt—low values indicate cobalt has reacted with tungsten to form non-magnetic eta phase (carbon deficiency), while high values can indicate free carbon diluting the magnetic phase. Coercivity measures resistance to demagnetization, which correlates with WC grain size—fine grains pin magnetic domain walls, increasing coercivity. Low coercivity means grains grew during sintering. High coercivity in a fine-grain grade is normal; in a coarse-grain grade it might indicate incomplete sintering. Every batch should be tested for both magnetic saturation and coercivity before shipping.

Dimensional control in powder metallurgy ultimately comes down to shrinkage control, and shrinkage is determined by the ratio of green density to sintered density. Higher green density means less shrinkage (part ends up larger); lower green density means more shrinkage (part ends up smaller). If all parts are consistently off-spec in the same direction, adjust pressing parameters. Variable dimensions within a batch indicate inconsistent green density from fill weight variation—powder feed systems that don't deliver consistent fill amounts cause this. Lot-to-lot variation is trickier: different powder lots may have slightly different particle size distributions or compositions that affect shrinkage even at identical green density. Tracking shrinkage per powder lot and adjusting die sizing accordingly is essential for tight tolerance work.

This sensitivity analysis shows how process variation translates to dimensional variation. Fill weight is the most critical: a ±1% variation in fill weight directly causes approximately ±0.3% variation in final dimensions, which is 30 μm on a 10mm feature. Green density variation of ±2% (which might result from inconsistent pressing conditions) causes ±0.6% dimensional variation. Sintering temperature and time have smaller effects but still matter for precision work. The key insight: dimensional consistency starts at powder filling, not at sintering. A well-controlled sintering process cannot compensate for inconsistent pressing. For tight tolerance work, weigh every fill and track the data—trends in fill weight predict dimensional trends.

Systematic troubleshooting requires isolating variables before changing anything. This table structures the investigation: what changed between good production and the defective batch? If the powder lot changed, that's the first suspect—test the new lot against a retained sample from the last good lot. Operator changes can introduce procedural variations even with trained personnel. Equipment changes include maintenance, calibration adjustments, or using backup equipment. Environmental changes (temperature, humidity, power quality) are often overlooked but can affect multiple process steps. If nothing changed but defects appeared, look for gradual drift in parameters that crossed a threshold. Answer all five questions before proposing corrective actions.

Before changing anything, collect complete data from the defective batch. This table lists the minimum data for meaningful troubleshooting. Powder data links to supplier quality; pressing data reveals compaction conditions; green part data shows what went into the furnace; sintering data documents thermal processing; sintered part data shows what came out; property data confirms final quality. Missing any of these stages creates blind spots in your analysis. Many troubleshooting failures occur because someone adjusted sintering temperature when the root cause was in pressing—complete data prevents this. If data wasn't recorded during production, you may need to run a repeat trial with full instrumentation to capture what actually happened.

This comparison template systematically identifies what's different between good and bad production. Fill out the "Current" column from defective batch data and "Last Good" from your baseline (ideally a successful lot made with the same powder lot). Calculate differences and flag anything exceeding the significance threshold. The thresholds are based on typical process sensitivity: >1% fill weight variation affects dimensions; >2% green density variation affects both dimensions and properties; >5°C temperature variation affects grain size and density. Significant differences become your investigation priorities. If nothing exceeds thresholds, look for combinations of small variations that together cross a process boundary, or expand the comparison to include more parameters.