By filling head
Identifies head-specific bias, spread, cut-off and maintenance issues.
Evidence and acceptance
How do you read a volumetric filling trial report?
Read a volumetric filling trial report by checking the test conditions first, then the individual results, distance from target, variation, filling-head pattern and start-stop states. A result is useful only when it can be connected to a defined product, pack, machine configuration and measurement method.
Individual resultsBias and spreadHead identityOperating state
Direct answer
Read the conditions before reading the headline result.
Confirm the product batch and temperature, fill target, container, machine configuration, feed arrangement, head identity, sample sequence and measurement method. Then assess how close the fills are to target and how widely they vary. An average without its conditions and individual values is not a complete acceptance record.
- Keep raw measurements available.
- Separate results by head and operating state.
- Record every setting needed to reproduce the test.
- Explain exclusions, rejected samples and interventions.
Useful measures
Which measures are useful in a volumetric filling trial?
The individual results are the foundation. The mean shows the centre of the measured sample; signed error shows whether that centre is high or low relative to target; range shows the distance between the highest and lowest values; and sample standard deviation describes spread. Coefficient of variation can help compare relative spread across different target quantities when used carefully.
| Measure | Question it helps answer | Important limitation |
|---|---|---|
| Mean | Where is the sample centred? | Can hide individual extremes and opposing head biases |
| Signed error | Is the centre high or low versus target? | Depends on a correct target and measurement method |
| Range | How far apart are the observed extremes? | Strongly influenced by sample size |
| Sample standard deviation | How dispersed are results around the mean? | Needs enough representative observations |
| Coefficient of variation | What is spread relative to the mean? | Can mislead when means are very small or conditions differ |
Separate the data
Why should results be separated by filling head and operating state?
Separating results by head reveals persistent mechanical or recipe differences. Separating by operating state reveals start-up, restart, replenishment or blockage effects that steady running can hide. Combining every fill into one sample can produce an acceptable overall average while a particular head or transition condition remains unsuitable for production.
By time sequence
Shows drift, product change, warm-up, settling or feed depletion.
By operating state
Shows start-up, stop, restart, blockage and replenishment behaviour.
Transfer to production
What makes a trial result transferable to production?
A trial result is transferable when the production product, preparation, temperature, feed route, container, machine hardware, recipe, speed and measurement method can reproduce the tested condition. The report should define those boundaries and identify what remains to be verified during installation or site acceptance. A demonstration with substitute product or ideal manual feeding is useful for learning, but it is weaker production evidence.
Record the accepted process, not only the accepted number
The approval should connect the result to the machine configuration, product condition and operating method that created it. If one of those changes, the process should be reviewed and, where necessary, reverified.
Related guidance
Continue the specification in the right order.
Send the raw results and the conditions that created them.
Provide the target, individual measurements, sample order, head identity, product condition and machine settings so Lancing can interpret the evidence correctly.