A practical evaluation framework for casting and core-room maintenance teams
When a foundry core box begins producing cores with surface defects, sticking, blocked vent paths or inconsistent release, the problem may be more than normal wear. Resin and binder residues can accumulate around the cavities and vent openings, especially where access for brushes and scrapers is limited. The maintenance decision is not simply how to make a dirty mold look clean: it is how to remove the relevant contamination while preserving the geometry that makes repeatable cores possible.
At OLLITAL, we recommend evaluating dry ice blasting as a controlled maintenance process rather than as a universal replacement for every cleaning method. It can be a useful option for selected core boxes, permanent molds and die-casting tooling, but cleaning intensity, part temperature, surface treatment, ventilation and downstream inspection must all be confirmed on the real tooling. This guide explains what to inspect, where the method can help, what it cannot promise and how to structure a defensible trial before buying equipment.
Cold-box and shell-core processes leave different residue patterns. A cold-box core box may collect deposits from binder carryover, overspray, resin mist, sand fines and release products. Shell-core tooling may accumulate heat-affected binder residues on cavity edges and surfaces. The deposit often appears as a thin film at first and becomes more tenacious after repeated cycles. Worn seals, poor vent maintenance or changes in the material system can also increase sticking and cleaning frequency.
For a useful diagnosis, separate three things: loose sand that can be extracted, an adhered binder or release-agent layer that needs removal, and true damage such as vent erosion, chipped cavity edges or out-of-tolerance tooling. Cleaning addresses deposits; it cannot restore damaged steel, rebuild screen geometry or correct a defective sand/binder recipe. When a defect is caused primarily by process settings or tooling wear, cleaning alone may provide only temporary improvement.
Start with vent slots and screens, cavity corners, parting lines, ejector regions and local areas where sand is repeatedly retained. Compare affected cavities with less-problematic cavities on the same tooling, if possible. Photograph residue before removal and record whether the issue appears after a specific number of shots, after a binder change or only on selected core geometries. These observations become the baseline for any core box cleaning trial.
A dry ice blasting machine accelerates solid CO2 media with compressed air. The pellets strike the deposit and sublimate; they do not remain as conventional abrasive grit. This is attractive where extra blasting media would be difficult to remove from small cavities. Nevertheless, dislodged binder, sand fines, oil and coating particles remain physical waste. They must still be controlled and collected by the site’s extraction and housekeeping system.
Dry ice blasting is generally a lower-abrasion option than blasting with hard mineral grit, but “lower abrasion” must not be read as “zero risk.” High jet force, close standoff, a sensitive coating or a cold-shock-sensitive insert can still create an unacceptable outcome. Apply the manufacturer’s guidance and trial the proposed setting on a representative area before cleaning high-value production tooling. Do not direct the stream into delicate screens, seals, instruments or bearings without explicit approval.
Deeply blocked vents may require a separate inspection, controlled manual procedure or replacement of the insert. Dry ice particles and air do not guarantee a through-cleaned vent. The exit path must be confirmed, and cleaning should not force contaminated particles deeper into an inaccessible passage.
The trial area must be a controlled maintenance zone, not a production demonstration conducted around moving mechanisms. Stop and isolate motion, pressure, electric power and other hazardous energy according to the site’s lockout/tagout procedures. Coordinate with the core-room manager and the tooling owner before touching high-temperature tooling; an in-place method does not automatically mean it is permissible to clean while the process is running.
Dry ice releases carbon dioxide gas as it sublimates. Confirm adequate ventilation and site-appropriate CO2 monitoring; do not rely on the visible vapor cloud as an indication of gas concentration. Review confined or low-lying spaces, adjacent pits and extraction discharge routes. Control dust and dislodged binder according to the material safety information, particularly where deposits could contain irritating or combustible constituents. Operators also require eye/face protection, insulated gloves for pellet handling, hearing protection and suitable respiratory protection where the risk assessment calls for it.

Dry ice media handling near a maintenance area; keep pellets appropriately stored and prevent avoidable exposure before use. Illustrative scene.
Media consistency matters. A bag of pellets that has been open for a long period may contain broken fragments or clumps. Bridging, wet compressed air or unstable feed can cause pulsating cleaning rather than controlled removal. Stage only the quantity needed for the planned work period and maintain dry, suitable compressed air. The complete air system - compressor, dryer, filters, distribution line and hoses - must deliver the pressure and flow needed at the machine inlet while blasting.
Select a core box with a known, repeatable contamination problem. Agree on the exact surfaces to clean, the surfaces to protect and the inspection method before blasting begins. Define success using practical criteria: visual deposit removal, vent condition, tool-surface preservation, cleaning time and the quality of subsequent core samples. Avoid vague statements such as “completely clean” without photographs or measurable criteria.
Photograph the same cavity details under consistent lighting. Record tooling ID, core process, material or binder system, cycle history since last clean, residue location, pre-existing damage and any blocked vents. Note whether the tooling is warm or cooled and whether the job will be performed in-place or on a dedicated bench. A before/after comparison without a matching baseline can be misleading.
Use the nozzle and dry ice particle size approved for the actual machine. Begin with a modest, controllable media feed and jet intensity, then adjust only after checking removal and surface response. Keep the nozzle moving; avoid concentrating the stream on a screen edge, fine feature or coating. Record what setting changed and how the result differed. Do not publish a universal “correct pressure” for all core boxes because tooling materials, deposits and machine configurations vary.
Stop, isolate and inspect the area after the short test. Confirm that slots and screens are not bent, loosened or newly obstructed, and that the deposit has been removed rather than redistributed. Examine parting-line edges, precision inserts, cavity finish and seal interfaces. Collect released sand and binder residues according to site procedures. If damage or uncontrolled debris is observed, the trial should stop until the cleaning plan is revised.
A good-looking surface is only an intermediate result. Run representative cores under the approved production conditions and compare sticking, core integrity, vent-related defects, scrap, rework and the time until cleaning is needed again. A trial is successful only if the cleaning outcome is consistent and does not introduce a new quality or safety problem. Record the total time from machine isolation through safe restart, not only the seconds when the nozzle trigger was depressed.
We recommend giving potential suppliers the same trial conditions, then recording observed rather than advertised performance. The rows below can be copied into an internal trial form. Leave fields blank until real plant data exists.
| Trial item | Record / acceptance question |
|---|---|
| Tooling and residue | Tool ID, binder type, deposit location, time since previous clean |
| Site controls | Isolation complete; ventilation / CO2 monitoring / extraction approved |
| Media and air | Dry ice grade/size, compressor delivered air, stable inlet pressure |
| Cleaning setup | Nozzle, standoff strategy, feed setting, test-zone area |
| Surface/vent inspection | Photographs; no new defects; vent openings verified |
| Production evidence | Subsequent cores accepted; scrap/rework observations |
| Time and materials | Total downtime, operator hours, dry ice consumed, debris handling |
| Disposition | Pass / modify / reject against pre-agreed acceptance criteria |

Controlled cleaning trial on casting tooling with local extraction; trial settings and surface compatibility require site approval. Illustrative scene.
For foundry work, the key purchase decision is the combination of cleaning task, tool sensitivity, compressor capacity, access and duty cycle. A compact setup may suit periodic cleaning of a few core boxes. A heavier maintenance program may require longer operating periods, a selection of nozzles and more robust media logistics. The presence of multiple machines in a catalog does not establish that any one of them has been validated on the buyer’s specific tooling.
OLLITAL currently lists the OLT-B3+ for industrial mold oil and residue cleaning, including casting molds. The published reference information lists 2.5-3 mm granular media, adjustable ice delivery, and a 13 kg hopper. Treat those as website reference figures: actual delivered configuration, inlet-air requirement, accessories and cleaning result must be confirmed in a technical agreement and on the proposed application.
Review the OLLITAL dry ice blasting machine range and the OLT-B3+ industrial mold-cleaning model before sending a test requirement. For air delivery and conditioning, the compressed-air sizing guide explains why compressor nameplate power alone is not enough.
Ask a supplier which nozzles are suitable for broad flat surfaces versus narrow recesses, how the operator adjusts media flow, whether anti-bridging precautions are required, what routine wear items exist, and whether dry ice can be sourced in the approved size near your foundry. Confirm portability and hose reach against actual aisle, machine and lifting access. Never infer that a model can handle all thin screens, ceramic inserts or fragile coatings from a generic “mold cleaning” claim.
It is tempting to compare only the blast time with the minutes required for brushing. For a foundry, the better measure is the total approved cleaning cycle: isolate equipment, wait for permitted temperature if required, open access, set up extraction, clean, inspect, collect debris, reassemble, release isolation and restart. Include rework and scrap only when they can be attributed to the same baseline period.
A simple evaluation is: cost per accepted core after maintenance = (cleaning labor + dry ice + compressed-air cost + related consumables + attributable downtime + disposal or extraction work) divided by accepted cores produced over the agreed comparison period. If the process is being assessed as a downtime improvement, record production hours recovered separately; do not count the same value twice as both reduced maintenance cost and additional output without explaining the assumptions.
One shift of improved access may be valuable even if pellet cost is higher than a scraper, but that is a conclusion for the plant’s own trial data. OLLITAL does not present estimated savings as a guaranteed percentage for an untested foundry.
Sometimes, if the tooling and machine design permit safe access and the cleaning method is approved. Installed does not mean running. Stop the process, isolate relevant hazards and follow site safety and temperature limits. The required opening, protective guarding and extraction route should be confirmed before work.
No. Dry ice blasting may remove deposits at accessible vent faces, but a deep blockage, damaged insert or packed residue inside a passage can require another maintenance method. The important check is verified vent condition after cleaning, not just a better-looking surface.
No. The dry ice media turns to gas, but removed binder, sand, coating and oil remain. Local extraction and appropriate debris collection are often required. The process does not automatically remove the obligation to manage hazardous contaminants.
No universal compatibility can be assumed. Coatings, soft alloys, seals, sensors, brittle inserts and existing surface damage must be assessed. Test a small representative zone and inspect it before extending the cleaning area.
Send core-box photographs, tooling material and coating details, resin/binder residue description, vent geometry, available compressed air, cleaning frequency, operating environment and the desired maintenance window. Include your trial acceptance criteria if available. This allows the team to discuss a realistic cleaning configuration rather than an ungrounded model recommendation.
The most reliable foundry mold cleaning programs are built around repeatable decisions: why a core box needs cleaning, which zones may be treated, which deposits have to be removed, how surfaces and vents are inspected and how successful cores are verified after restart. Once these controls are documented, equipment selection becomes clearer and operators can repeat the result across shifts.
At OLLITAL, we recommend a small trial with representative tooling before committing to a broad change in maintenance practice. It is also worth asking suppliers about training, consumables, technical documentation and after-sales support; our dry ice blasting manufacturer evaluation guide provides a practical way to compare those items.
Have a core box with recurring binder buildup or blocked vent slots?
Share your mold photos, contamination type and maintenance requirements with our team. We can discuss trial conditions and a suitable configuration before you make an equipment decision: Contact OLLITAL.