Graphite Grid Moulds for Powder Sintering: Hundreds of Cells Arranged on One Plate – Why Is the Hard Part Not Cutting the Cells, but the Hundreds of Thin Walls?

XRD Graphite · Graphite Grid Moulds for Powder Sintering · Technical Column

Graphite Grid Moulds for Powder Sintering: Hundreds of Cells Arranged on One Plate – Why Is the Hard Part Not Cutting the Cells, but the Hundreds of Thin Walls?

The typical form of this component is a rectangular graphite block with rows and columns of cell holes arranged across the plate surface—the holes are small and dense, leaving only a thin wall between adjacent cells, with a border frame around the perimeter. It belongs to the same family as cavity plates with few, widely spaced cavities, but the account is different: when cavities are sparse, the difficulty is “making every cavity identical”; when cells number in the hundreds and the walls are reduced to a single thin layer, the difficulty shifts to “whether those hundreds of walls can be equally thin and equally neat.” This article covers three things: how it differs from the sparse-cavity approach, which accounts change together once the walls are thin, and several points worth specifying in advance on drawings and acceptance.

Written by: XRD Graphite · Technical Column

I. First, Positioning: How It Differs from the Sparse-Cavity Approach

Within the same family, there are at least two arrangements: one with few, widely spaced cavities that are large and deep, with relatively wide ribs between cavities; the other with many, densely packed cells that are small with thin walls, the whole surface resembling a regular grid. The dividing line between them is not “more cells or fewer cells,” but a single word—wall.

• Sparse-cavity approach: Cavities are large, ribs between cells are left thick, rigidity is built in, the tool has room to move during machining, and tolerance for error is greater.
• Dense-cell approach (the type discussed here): Cells are small and packed tightly, leaving only a thin wall between adjacent cells. Wall thickness is often not something the designer freely chooses, but is forced by product dimensions—it must enclose the forming space, conduct heat, and support itself.

This is what makes this type of component special: there is no allowance on the walls. In sparse-cavity parts, allowance is left on the ribs; in dense-cell parts, there is nowhere to leave allowance, and it must be preserved by the machining process. Therefore, almost every account on this component ultimately revolves around “the wall”—wall rigidity, wall uniformity, wall loss, wall cleaning.

It is also worth distinguishing its position from adjacent component types: sintering support plates and sintering discs only hold workpieces and do not hold material themselves; graphite boxes and graphite boats are a single large cavity holding a batch of material; plates with through-hole arrays use holes for venting or for rods to pass through. The type discussed here uses a densely distributed field of cell holes to separate many portions of material in one furnace, forming or supporting them simultaneously.

II. Once the Walls Are Thin, Four Accounts Change Together

Cutting the cells is not difficult; what is difficult is that once the walls are thin, the entire calculation method changes.

First, rigidity of a single wall: tool deflection and chatter marks go from occasional to routine. A thin wall has no support on either side. When the tool reaches it, the wall is soft. Milling force pushes it aside; the thinner the tool and the longer the overhang, the more pronounced the deflection. A deflection that occurs occasionally on a sparse-cavity part appears across the entire field on a dense-cell part—resulting in wall thickness that is thick on one side and thin on the other, uneven cell openings, and rippled cell bottoms. Such deviations often cannot be read by measuring a single cell with calipers; they require row-by-row sampling, templates, or dedicated gauges to detect.

Second, labour time follows cell count: the cost centre shifts from material to labour. Hundreds of cells mean hundreds of tool entries, hundreds of corner cleanings, and hundreds of dust-cleaning operations on one plate. Material cost is the price of one plate, but machining time rises with cell count. Therefore, the quotation focus for this type of component is labour, not material: one more cell on the drawing means one more cell of time at the machining end. In practice, clarifying at the quotation stage that “cell count directly determines labour time” is easier than explaining the quotation afterwards.

Third, low heat capacity and fast response: this has two sides. Once the wall is thin, its heat capacity is small and it follows temperature changes quickly, allowing the furnace rhythm to be tightened—this is the good side. On the other side, fast response also means it is more sensitive to local temperature fluctuations in the furnace—wherever more heat is supplied or atmosphere flows more slowly, it is reflected more quickly on the thin wall. Thick parts are “stable,” thin parts are “responsive”; stability and responsiveness each have their applications. Before selecting, clarify which one this process step needs more.

Fourth, effectively stricter: the same tolerance value represents a larger proportion on a thin wall. A wall thickness tolerance is very loose on a thick rib; on a thin wall, the same deviation represents a much larger proportion and is effectively much stricter. The same applies to hole spacing, wall thickness, and flatness—so tolerances on the drawing should be matched to the scale of the feature itself. Applying a single accuracy grade to the entire component commonly results in places that should be strict not being strict enough, and places that need not be strict wasting labour time.

The relationship among these four: rigidity determines whether walls are machined uniformly, labour time determines whether the job is worth doing, heat capacity determines how responsive it is in the furnace, and equivalent tolerance determines whether the drawing is written to the point. All four matters rest on a single thin wall—this is the real difference from sparse-cavity parts.

III. At the Machining End, Where Does Labour Time Concentrate?

From a blank to a grid plate ready for the furnace, the process roughly follows these steps: determine plate format, thickness, and cell layout; grind both large faces flat, establishing flatness as the datum; rough-cut cells, leaving finishing allowance; finish-machine cell shape and cell bottom, clean corners; treat cell openings and border; verify hole spacing and wall thickness piece by piece, verify flatness for the entire plate, and blow dust clean. The process looks conventional, but labour time concentrates in the following areas.

• Datums first. Densely packed cell holes require several setups and several flips. Every datum change shifts cell positions. The reliable approach is to first produce a set of datum faces and datum holes, then measure all cell positions from them, rather than finding alignment separately in each operation—the more cells, the greater the cost of inconsistent datums.
• Tool wear accumulates within a single plate. Hundreds of cells on one plate are usually cut with the same tool, and the tool wears continuously during the run. Wear directly translates into drift in cell opening dimensions and cell depth: cells at the beginning are slightly larger, cells at the end slightly smaller. Measured individually, all are within the tolerance band; placed together on the plate, they form a trend line. A workable approach is to write tool change timing into the process, segment by tool life, leave re-inspection points between segments, and if necessary divide cells into groups calibrated by small-batch first articles, rather than waiting until the entire plate is finished to sort.
• Chipping is interconnected. Both sides of a thin wall are free edges. If one location chips, both adjacent cells are affected—this differs from sparse-cavity parts, where a chip generally affects only one cavity. Graphite is brittle; cell openings and corner-cleaning locations are prone to chipping. Feed method, cell-cutting sequence, and tool sharpness must all be determined in advance. Chipping caused by improper clamping often occurs near the end of finishing, and the cost is visible.
• Clamping and springback. Once the plate is hollowed into a grid, its rigidity drops. Excessive clamping force makes it spring; after machining, releasing the clamp lets it spring back, and flatness is lost. Support points and pressure points should fall on the border or at wall intersections, separating force application from cell holes.
• Dust cleaning is a fixed major item. Cells are small and numerous; places where compressed air cannot reach are hidden risks. Residual powder left in cells will fall onto green bodies after furnace loading, directly contaminating appearance-sensitive products. For multi-cavity components, dust-cleaning workload follows cell count. Blowing clean cell by cell, with verification where necessary, must be realistically included in labour time.
• Inspection conventions should be agreed first. How hole spacing is measured, how many points are measured, and how much measuring force is controlled should be agreed in advance rather than argued afterwards—graphite is relatively soft, and a slightly heavier clamp with calipers produces a different reading.

IV. Several Points Worth Specifying First on Drawings

• Array marking and datums. For hundreds of holes, whether the marking is “array marking + which edge or hole is the datum” or sequential accumulation makes a significant difference. Sequential accumulation adds the deviation of each cell to the next; by the end of a row, the difference between the two ends becomes apparent. Measuring from a datum in one step is the proper method for an array.
• Uniform wall thickness. The wall is the core quantity of this type of component. In addition to hole diameter and hole spacing, the drawing should separately specify “uniform wall thickness” and note which section the measurement points fall on and how many locations are measured—without this, machining and inspection can only follow their own interpretations.
• Whether cell holes are through or blind. These two types serve different purposes: through cell holes are used for venting, heat transfer, rod passage, or material passage; blind cells are cavities for holding material, one portion of powder per cell. Both may coexist on the same plate. Which type it is must be stated on the drawing, not left for the machining end to guess.
• Which face flatness applies to. Flatness over the entire plate and flatness of a single cell bottom are two different things. Once the plate is hollowed into a grid, residual machining stresses and hot-state temperature gradients make it want to bow; once the plate bows, cells at different positions on the same plate have different actual depths. With the same powder load, parts come out thick on one side and thin on the other. Therefore, the drawing should specify not a single value but “which face and which area this value applies to.”
• Treatment of cell openings and corners. Chamfering or slightly rounding cell openings reduces the chance of chipping during tool entry and exit, leaves no dead corners during cell cleaning, and prevents scratching material during loading and unloading. The same applies to the transition between cell bottom and cell wall—sharp corners are stress concentration points in graphite and hiding places for residual powder.
• Border. After cells are packed to the edge, the border left around the plate bears all the weight and rigidity. Its width is determined by lifting points and support points, not left as random leftover material.

V. Where Are They Mainly Used?

Wherever a process requires “separating portions of material by cell and sending them into the furnace together,” this type of grid plate has a role. For a specific piece of equipment, the plate size, cell layout, cell shape, and wall thickness must all be determined based on product drawings, powder characteristics, shrinkage behaviour, and furnace loading method. This type of component basically does not offer universal values.

VI. Operating Boundaries: Four Things to Watch for Thin-Walled Components

Graphite is not afraid of heat per se—it is the combination of heat and oxygen that poses the risk. This also applies to grid plates, and thin-walled components are especially sensitive to it—which is also why they are mostly used in vacuum furnaces and protective atmosphere furnaces.

• Oxidative loss. Graphite undergoes oxidative loss in oxidising atmospheres such as air at sustained elevated temperatures. Thin walls have a large specific surface area, and locations such as cell openings and corners are often damaged first. Because the wall is already thin, the same layer of loss represents a larger proportion; once a cell opening enlarges and dulls, the dimensions and weight of the output drift first. Vacuum or inert gas protection avoids this problem. If use in an oxygen-containing environment is required, loss must be anticipated and corresponding protection provided.
• Blockage and scaling. Binder volatiles and powder residue from sintering gradually deposit on cell walls. Small cells are more prone to blockage than large cavities—block one cell and one usable cell is lost, reducing loading capacity. If cleaned, cell dimensions can be maintained; if not, the cell is effectively shrinking. Periodic cell cleaning by furnace run is recommended, with cell opening dimensions re-inspected when necessary.
• Impact. Thin walls are the vulnerable locations on this type of component. A chipped cell opening during loading, unloading, or cell cleaning affects the output of both adjacent cells and reduces the number of usable cells. Cell openings, inner corners of the border, and plate corners are key locations for impact protection.
• Fit and thermal expansion. Reasonable clearances must be left between the plate and the material frame, backing plate, and pressing tool. Thin walls have low heat capacity and respond quickly; if clearance is too small, they are more likely to bind in the hot state and create local stress. If too large, cell positions and green body positions drift. Thermal expansion differences between mating components must also be calculated together with dimensions after heating.

Then there is its consumable nature. Cells gradually lose accuracy after repeated cleaning, thermal cycling, and demoulding friction, with thin walls deteriorating first. Rather than holding out until the entire batch output begins to fluctuate, it is better to inspect periodically as a consumable and replace at the appropriate time—the cost of one plate is usually lighter than the batch of green bodies it affects.

VII. About XRD Graphite’s Graphite Grid Moulds for Powder Sintering

With 30 years of experience in graphite grid moulds for powder sintering, XRD Graphite has accumulated extensive process expertise in milling, corner cleaning, and dust removal for thin-walled, densely packed cell holes. We machine precisely to customer drawings and strictly control critical tolerances such as hole diameter and hole spacing, wall thickness uniformity, cell depth, and overall plate flatness. Our products offer relatively long service life under normal use and maintenance and have gained recognition from many industry clients. The factory is located in Baofeng, Pingdingshan, Henan, with both graphite material production capacity and machining lines. Material preparation and machining for such components can be completed in-house.

We can custom-manufacture graphite grid moulds for powder sintering and various densely packed cell-hole graphite tooling components, and support custom orders from drawings, samples, and small-batch trials. A more reliable approach is to first clarify the product drawing, powder characteristics and shrinkage behaviour, furnace loading method, and atmosphere conditions, then match the graphite grade and cell shape plan accordingly. If drawings or operating conditions are not yet clear, single-piece or small-batch verification first, running through cell depth, wall thickness, and flatness once, then scaling up the full batch, keeps trial-and-error cost relatively controllable.

Final Thoughts

A graphite grid mould for powder sintering may look like a grid on a plate, but it carries the account of “the wall”: cells are easy to cut, walls are hard to preserve. Once the wall is thin, four accounts change their calculation method together—rigidity of a single wall, labour time within the plate, response in the furnace, and the actual tightness of tolerances. This is the true dividing line from the sparse-cavity approach. Half of this account falls at the machining end, through unified datums, segmented re-inspection, light cutting, and cell-by-cell dust cleaning; the other half falls on the drawing, clarifying datums, wall thickness, measurement methods, and flatness conventions first. In actual procurement, presenting the product drawing, powder characteristics, and furnace loading method first, and then discussing cell shape, grade, and inspection conventions, is the more reliable path. For specific operating conditions, confirmation based on actual parameters and drawings is still recommended.