inquiry
Leave A Message
If you are interested in our products and want to know more details,please leave a message here,we will reply you as soon as we can.
submit
Home

parting line stub on castings

parting line stub on castings

  • Parting-Line Flash on Castings: Why an Average Grinding Path Leaves Stubs or Gouges the Face
    Nov 11, 2026
    Parting-Line Flash on Castings: Why an Average Grinding Path Leaves Stubs or Gouges the Face   Outline 1. What inspectors flag is often the parting line, not the big riser. 2. Flash, fins, and riser stubs are different cuts. One parameter set will not cover all three. 3. Thickness variation, mismatch, and fin path have to be separated before programming. 4. A path set to average thickness leaves a stub on the thick side and cuts parent metal on the thin side. 5. Tools and contact force chosen only to “get it off” hit seal faces and thin walls first. 6. Faces the parting-line path must not sweep across. 7. Hand the line to a machine only when the batch is stable and the flash can be described. 8. What to measure and photograph before the first cycle. Castings sent back from the inspection table are not always the ones with a riser left standing. More often an inspector runs a finger along the parting line and stops: a stub here, a scooped corner there. Grinders know the difference. Open faces and outside diameters are easy to lean a grinder into. A parting line that dives behind a boss, wraps a bolt ear, or disappears between two ribs is not. The wheel will not sit, and a small twitch leaves a pit. The difficulty is not a vague “complex process.” The line itself is uneven. The same pattern and the same heat of iron can throw a fin that is almost flush on one stretch and two or three millimeters proud on the next. Sand castings add mismatch, so the two sides of the parting face are not one plane. A robot path set to an average thickness then does one of two things: the thick stretch keeps a root, and the thin stretch loses parent metal. A fin is not the same cut as a riser stub Shops call all of this grinding, and programs are often written as one continuous path. The force case is not the same. A riser root wants rigidity. On a large iron stub, once cutting force builds, the arm and spindle have to hold. If they yield, the root stays as a step. Wheel or mill engagement is deep, the path is short, and the clamp has to be checked against that force. A parting line is the opposite. It is thin, long, and uneven, and it often runs next to a seal face, a machining datum, or a cosmetic face the customer has marked off-limits. The risk is not failing to cut. It is cutting too much. Aluminum die castings make this obvious. The fin is often paper-thin. A rigid mill pass drops the fin and leaves a step on the parting face. That step is what the OEM feels. Iron fins can carry sand. Load on the wheel jumps, and a fixed feed will dwell or dig at the inclusion. Steel parting lines with burnt-on sand and scale eat a thin cutoff wheel faster than clean flash. If the wheel is not changed, the path is still there and the cutting edge has moved back, so the stub returns. One tool and one feed can finish the cycle. They will not necessarily finish the part. Thickness, mismatch, and path direction are separate problems Variation gets reduced, too quickly, to whole-part offset compensation. A displacement sensor reads the difference against the fixture, and the controller shifts the path. That helps when a gate has moved as a block, or the whole mold is shifted. Flash along a parting line is not always that kind of shift. A thick stretch and a thin stretch are local. A sensor watching one datum computes one correction for the part. It does not remove thickness change along the line. The path still carries one stock allowance, and the thick-stub / thin-gouge pair remains. Mismatch needs its own look. Once the cope and drag are offset, the parting line is a step, not the curve on the model. A path glued to the CAD parting face cuts parent metal on the high side and misses the fin on the low side. If the mismatch is outside the casting tolerance, it belongs back in the foundry, not in the grinding cell. If it is inside the allowance, the path has to be split: relief on the high side, cleanup on the low side. Averaging the step does neither. Direction has to come from the part, not only the print. The drawing shows a closed curve. The casting may break that curve with ejector marks, overflow, or a vent wafer. When the vent is tied into the fin, a grinder flicks it off in the same motion. A robot with no separate segment leaves a sharp ear. Assembly cuts hands on it, and the customer calls the part unfettled. Why a fixed path fails in the same corner A few trial parts pass, then the batch gouges the same corner. Repeatability is rarely the cause. Good repeatability only means the tool arrives at the same point. If that point was taught on a thick sample and the next lot is thin, the edge is still there and the parent metal goes. The reverse happens too. The sample fin was heavy, the path was backed off, then the pattern was repaired and the fin got thin. The machine still runs the old path. The motion looks complete. The finger still finds a root. Pattern repair, pattern change, and sand compactness all move the fin. If the path is not checked again, automation has only made one defect stable. Compliant contact helps and does not cover everything. Current, position, and torque loops let the tool yield over high and low spots. That suits faces that must not be cut hard. If the parting line sits on a seal band, too much force still lowers the band. Too little force leaves the thick fin. The setting belongs to that stretch of face, not to the whole line as one contact value. Match the tool and the force to the face On open iron or steel parting lines, a thin cutoff wheel is fast and also the tool most likely to cut a crescent at a corner. Corners and ear roots are steadier with a small mounted point or rotary burr, at lower speed, than with a cutoff wheel forced around the bend. If the customer wants a hand feel after the root is gone, a flap wheel can knock the scratch pattern down. A flap wheel is not the roughing tool. On a heavy fin it only polishes, and the root stays. Aluminum should not inherit the iron wheel habit. A disc cutter or small mill can break a thin fin, then a flap disc or brush can dress the edge. That usually raises less of a step than one rigid lap around the part. A mill larger than the opening never reaches an internal fin. Spindle power is irrelevant until the tool can get there. If it cannot, flip the part in the fixture or leave that stretch manual. Do not write it up as fully automatic. Thin walls bring a clamping issue. A parting line on a flange, with a clamp near the fin, bows the part before the tool arrives. The path cuts a deflected face. When the clamp opens, the pit appears. Check the clamp distance before talking about compensation. Some faces must stay outside the path Pump, valve, and hydraulic seal bands are often checked with a straightedge or spotting compound. The parting line runs at the edge of the band, and a grinder in a hurry widens the band. A robot path that overshoots by a couple of millimeters to clear the root will scrap the band across the batch. That face needs a keep-out. Root cleanup stops outside the band. On the band, only a proven light pass, or no pass at all. The same rule covers machined datums, bearing seats, and tapped-hole mouths. Whether grinding sits before or after machining sets how deep the parting line may go. Grind first, and a small root may disappear in machining. Machine first, and a wheel scratch on the machined face scraps the lot. If that order is not in the process, the equipment supplier and the foundry will argue past each other. Cosmetic parts also care about scratch direction. Across the line and along the line do not reflect the same way. Export covers are often rejected for uneven hand feel, not for a dimension. Match path direction to the customer’s inspection method on the sample, not after the batch is out. Not every parting line belongs on a robot High repeat work, a stable pattern, and flash that a few samples can describe are the fit. A dual station lets loading run while a long parting-line path is cutting. If the pattern changes daily and nobody records fin shape, programming and rechecking will outlast the grinding. Prototypes and patterns still being cut are better left to a skilled hand on the parting line, with the machine taking risers only. Once fin thickness and mismatch sit in a band for several lots, freeze the path. The opposite priority is also real. On a running job where people only rework the parting line, automate that line before adding a machine to an outside diameter that is already stable. Iron production parts can use a rigid spindle and a split path. Thin aluminum fins depend more on tool diameter, a flip, and a light dress. Steel has to put wheel wear and wheel changes into the cycle time, not only the cut. Those differences matter more than “add a robot.” Anhui Neview Automatic splits iron, aluminum, and steel across the DR, MR, and GR series. On a parting-line job the checks are still whether the tool reaches, whether keep-out zones are drawn, and whether local thickness is being forced into a whole-part offset. What to measure before the first cycle Take three to five castings in a row, not the best one on the pallet. Measure fin thickness along the line: thick stretch, thin stretch, corner. Write mismatch as a height step across the parting face, not as “a bit of shift.” Circle seal bands, machined faces, and thread mouths, and mark what may be touched. Photograph vents and overflows that join the fin, and mark the cut-off point. Set wheel or cutter size for that stretch. A corner tool change belongs in the cycle time, or the floor will force the same wheel around the bend to avoid a change. Note how far the clamp sits from the line, and check the flange with a straightedge after clamping. Do not accept a trial because the motion finished. Run a batch on the same path, feel the thick and thin stretches, and spot the seal band. A stub on the thick side means local depth or a tool change, not a deeper pass on the whole line. A gouge on the thin side means a lift or a light pass, not a slower feed dressed up as a fix. After pattern repair or a sand change, remeasure those three points. If they have not moved, leave the path. If they have, and the old program still runs, the stub and the gouge come back in the same corner. Whether a parting line is worth automating depends on whether it can be described. Flash that cannot be described will be cut wrong, steadily. Flash that can be described, with keep-outs drawn and a tool that actually reaches, is a better candidate for the machine than a broad face already easy to grind by hand.
    Read More

leave a message

leave a message
If you are interested in our products and want to know more details,please leave a message here,we will reply you as soon as we can.
submit
Contact Us: judy@ahxjj.cn

home

products

WhatsApp

Contact Us