Diagnosing the Causes of Molybdenum Crucible Cracking

Replacing a cracked crucible with a new one straight away often means the same failure comes back in the next campaign. The position, direction and fracture-surface morphology of a crack record how the load was applied and how the failure progressed. Reading that evidence before discussing material or operating practice makes the investigation far more efficient. The inspection scope and the sampling method for the failed part should also be agreed in advance, so that the original fracture surface is not destroyed during handling and cutting.

Read the Fracture Surface and Crack Direction First

A through crack running across the wall thickness is mostly related to thermal stress or mechanical constraint; a fine crack that extends only along the surface may be related to machining damage or localized oxidation. Whether the fracture is intergranular or transgranular indicates the temperature range of the failure and the distribution of impurities.

Avoid secondary damage while examining the part: photograph and record the crack position and length before cutting any specimen. Where necessary, section the part metallographically to see whether the crack tip shows oxidation tinting or second-phase agglomeration. If the fracture surface is already clearly oxidized, the crack opened at high temperature rather than from mechanical damage after the furnace was shut down.

Material Verification

Check the batch number, purity, density and grain size. Low density means a high pore content, and a crack links up easily along pores; a coarse grain size lowers toughness and weakens resistance to thermal shock. On purity, pay attention to interstitial impurities: high oxygen and nitrogen contents form brittle phases at the grain boundaries.

Material verification cannot rely on the mill certificate alone; where necessary, take samples from the same billet batch and re-test them. If several parts from one batch crack at the same position, a material consistency issue deserves to be suspected first. When abnormal inclusions or clustered pores appear near the fracture surface, take samples from that area for laboratory analysis and check the sintering records of the same billet batch.

Structure and Charging

Whether an adequate fillet is provided where the wall thickness changes, whether the support pedestal contact is flat and free of point contact, whether the charge exceeds the design height, and whether lifting and transfer caused impact or forced deformation: all of these belong to the structural line. Thin-walled crucibles are especially sensitive to point contact, where local stress turns directly into a crack origin. The contact face between pedestal and crucible should conform, avoiding three-point support or local bridging; forced alignment during assembly leaves residual stress in the crucible, which adds to thermal stress on heat-up, and the crack then usually starts near the contact point.

The match between charge height and crucible volume should be written into the drawing and the work instruction, so that the shop floor does not charge by habit. Operating records help reconstruct the event: charge weight, heating rate, holding time, whether the furnace ran empty, and whether it was quench cooled. If cracks appear after the cooling stage, the cooling rate and any positional constraint should be reviewed first.

Re-checking the Thermal Field

An asymmetric thermal field raises the temperature of a local region of the crucible; under prolonged exposure, creep and thermal stress concentrate in one fixed azimuth and the cracks then follow a regular pattern. Check the relative position of the heating element, the heat shield and the crucible, confirm that the crucible is centered and that the heat shield is not distorted. Distortion of the heating element, displacement of the heat shield and off-centering of the crucible all push the thermal field away from its design; measuring the gap between the crucible outer wall and the heating element at regular intervals reveals the change before cracking rather than after it.

Mapping the crack azimuth on the failed part to its azimuth in the furnace and comparing it with the temperature record often reveals the repetition. Wall-thickness measurement should cover both the region near the crack and regions remote from it; subtracting these readings from the original wall thickness at commissioning gives the amount of local thinning.

Record Comparison and Attribution

A single cracked part points to operating practice, a batch of cracked parts points to material and thermal field, and repeated cracking at the same azimuth points to structural constraint. This rough guide narrows the search but does not replace data. Each crucible should have a commissioning date, charge records, maintenance records and a failure description, so that batches can be compared. The value of records lies in their comparability: if crucibles, charges and heating curves from different batches are mixed together, the variables cannot be separated. Filing by furnace position and by batch allows comparable records to be retrieved quickly when a failure occurs.

When attributing cause, separate the direct cause from the contributing factor. Porosity in the material may be the direct cause while a fast heating rate is the contributing factor; point contact in the structure may be the direct cause while an overloaded charge is the contributing factor. Change only one of them and the problem may still recur.

Sampling and Submission of the Failed Part

The first step in failure analysis is to preserve the evidence. After the crucible is taken out, avoid hammering, cutting and re-installation; first mark the crack position with a marker pen and photograph it. Where sectioning is required, the specimen should cross the crack tip so that the original fracture morphology is retained.

Select the laboratory tests according to the suspected mechanism: composition and density, microstructure, fracture-surface scanning, wall-thickness measurement. If atmosphere or melt interaction is suspected, attach records of the service environment and samples of residue. State the operating conditions clearly on the test request so that the laboratory can give a targeted judgment. Avoid weld seams and heat-affected zones when sampling unless those areas are suspected crack origins; for batch failures, take unused parts as controls as well. Density and microstructure data from the control parts help separate material defects from service damage.

Putting Improvements into Practice

The conclusions must be turned into executable actions: adjust the pedestal, revise the heating curve, limit the charge height or switch to another batch. The actions should name an owner and a verification method, and be reviewed after the next batch is commissioned. Recording only the crack length without changing the conditions makes repeat failure hard to avoid. After the improvements are implemented, follow at least three production batches before judging whether they worked. For parts that failed repeatedly, also verify that installed dimensions match the drawing, and file the charge records together with the operating tickets.

For a supplier, failure analysis is also an input to process improvement. Correlating fracture morphology, density and sintering records shows whether raw material, forming or sintering is under proper control. Turning every failure into an improvement record is how crucible life accumulates.