On this page
How can a buyer compare and accept gravimetric blender dosing accuracy without relying on an ambiguous percentage claim?
Define what the percentage is relative to, which material and dose size are being tested, and whether the limit applies to individual batches or a cumulative run. Request component-level records, agreed low-dose and refill trials, and separate evidence for finished-part requirements. Set acceptance limits with the material and equipment suppliers before the trial.
Decision checklist
- State the recipe denominator and error formula explicitly, including how regrind is treated.
- Identify batch blending versus continuous loss-in-weight feeding before comparing specifications.
- Test the intended materials and agreed minimum dose, not only a convenient high-throughput recipe.
- Retain batch or interval data so a cumulative average cannot hide unacceptable excursions.
- Define startup, refill, restart and material-change conditions with the supplier.
- Keep dosing acceptance distinct from molded-part appearance and performance approval.
Make the percentage claim comparable
A quotation stating 'accuracy within 1%' is incomplete unless the supplier explains 1% of what and over which observation period. Request the recipe basis, target component mass, measured component mass and error calculation. Where additives are described relative to a base-resin quantity rather than the total blend, obtain an explicit conversion before comparing offers.
Maguire's historical technical note distinguishes relative target error from changes in blend percentage and discusses the influence of dose size and averaging period. Use that discussion to frame questions, not as a current performance guarantee for a quoted machine or a universal industry acceptance limit.
For an illustrative fixed-total batch of 2,000 g, a 1.00% additive target is 20 g. If the actual batch contains 21 g additive and 1,979 g other material, the additive fraction is 1.05%. That is an increase of 0.05 percentage point, but a 5% relative error against the 20 g target. This is arithmetic only, not a recommended tolerance or an observed test result.
Separate batch blending from continuous feeding
Conair distinguishes batch gravimetric blending from continuous loss-in-weight feeding. Both use weight, but a batch record and a time-based feed-rate record do not describe the same acceptance test. Identify the supplied architecture before specifying the evidence.
For a batch blender, ask how component targets and actual weights are recorded for each batch. For a continuous feeder, agree on the measurement interval and how its output is coordinated with the process demand. Do not compare a long-run average from one system with a single-event limit from another.
Build the trial around the material and lowest intended dose
The proposed table below is a buyer-side test-planning framework. The supplier and responsible process or quality engineer should approve the actual procedures and limits. Agree on material grades and lots, the intended recipe range, batch size or demand rate, and the supporting weighing method before scheduling factory acceptance.
| Trial condition | Record to request | Question the evidence should answer |
|---|---|---|
| Normal production recipe | Material IDs, recipe basis, component targets and actual quantities | Does the offered configuration meet the agreed limits under intended operating conditions? |
| Lowest intended additive dose | Smallest target mass or feed rate and individual observations | Is low-dose performance demonstrated rather than inferred from a larger dose? |
| Regrind-containing recipe | Regrind identification and documented treatment of every percentage | Are supplier and buyer calculating the same formulation? |
| Normal refill event | Time-marked refill and dosing records using the approved procedure | What happens during refill and how is affected output identified? |
| Startup or approved restart | Event log, stabilization rule and material disposition | Which output is included in acceptance and which requires segregation? |
| Material or color change | Cleaning responsibility, recipe revision and sample identity | Can the next run be traced without implying that weighing proves cleanliness? |
Ask for raw evidence before accepting a summary
Request a record format that links each observation to the equipment ID, controller configuration, recipe revision, material lot and trial condition. Ask for the supplier's weighing-verification record and the identification of any independent reference equipment used. Display resolution alone should not be treated as evidence of acceptable dosing performance.
Define in advance whether acceptance applies to each batch, a specified collection interval, the cumulative quantity, or more than one measure. Retain both the relevant individual observations and the agreed summary. Record excluded startup material and invalid observations with reasons rather than deleting them to improve the average.
Where a demonstration uses substitute material, label the result as provisional for the actual recipe. A sensible purchasing decision may be to reserve final acceptance until an agreed production-material trial is completed, instead of treating the demonstration as a guarantee.
Do not use dosing accuracy as a substitute for part approval
Correctly recorded ingredient quantities do not by themselves establish color uniformity, dispersion, appearance or the required finished-part properties. Define separate material and part acceptance criteria with the responsible suppliers and quality team. Keep sample identity connected to the dosing record so that an issue can be investigated without guessing which recipe produced it.
Similarly, a visually acceptable short run does not establish the accuracy of the weighing system. Treat equipment dosing evidence and molded-part evidence as complementary decisions. Agree on who owns each decision and how a discrepancy will be resolved.
If a project includes machine process acceptance as well as auxiliaries, retain separate records for temperature-control evidence rather than merging every result into one 'cell passed' statement.
Write a quotation-ready acceptance clause
A practical starting clause is: 'Supplier and buyer will approve the material list, recipe calculation basis, operating range, trial conditions, data fields and acceptance limits before FAT. The handover will include the agreed records and all open exceptions. Any untested condition will be identified explicitly.' Adapt this with the project team; it is a scope prompt, not legal advice.
For a GEERPOWER project review, send the resin and additive grades, intended regrind use, recipe range, machine demand, material-handling layout and destination. State whether the blender is supplied by the buyer, the auxiliary vendor or the cell integrator. Request written responsibility for installation, acceptance, training and support.
These recommendations do not promise a dosing accuracy, throughput, scrap reduction or return on investment. Those commitments require an identified equipment configuration and an agreed evidence-backed quotation.
Common buyer questions
Is a gravimetric blender automatically acceptable for every low-dose additive?
No. Ask the supplier to demonstrate the intended material and minimum target quantity with the offered dispensing configuration. A result at a larger dose or with substitute pellets should not be assumed to cover your application.
Can total material consumption prove that every batch met the limit?
No. A cumulative result can be acceptable while individual batches differ. Define whether the requirement is batch-specific, cumulative or both, then retain the records needed for that decision.
Should startup and refill data be removed from the acceptance report?
Not silently. Agree on the operating conditions included in acceptance, identify events in the record and document the disposition of affected material. Any exclusions need an agreed reason.
Does an accurate ingredient ratio guarantee the required molded-part color?
No. Dosing evidence does not replace the agreed finished-part appearance or performance evaluation. Link part samples to the recipe and dosing records and approve those requirements separately.
Sources and references used for this guide
This guide supports early project planning. Final machine selection and commercial configuration must be confirmed from the actual part, mold, material, factory utilities and production target.
