The specification sheet gives a filtration accuracy of ≤1 μm, and many people read that as meaning every particle larger than 1 μm will be held back. That does not match reality. The nominal accuracy is a rated value given under specific conditions: the material and structure of the filter medium, the rated flow and allowable differential pressure of the element, the particle distribution used for the test, and whether the test is single-pass or multi-pass. When these conditions change, the same element behaves differently on different sites.
The nominal accuracy describes the medium, not the whole machine
Two levels have to be separated first. The nominal accuracy of a filter element is a property of the medium: the fibre diameter, the weave and the surface treatment of the medium determine its ability to retain particles of different sizes. The filtration accuracy of the machine as a whole is a different matter, determined jointly by the specification, the stage count, the sealing condition and the bypass situation. The filtration accuracy of the ZYD-100 as a whole is ≤1 μm and the corresponding cleanliness is NAS 1638 class 5, which is the result of the machine correctly assembled, correctly filtered and run at the correct flow rate.
Among the commonly used element specifications, the coarse filter elements are 150-250-50 and 110-250-50, the second stage elements are 110-400-20, and the fine filter elements come in two types, 110-600-5 and 150-500-5. The standard configuration of the ZYD-100 is 5 elements of 150-500-5 in parallel. If the fine element is changed for one of a different specification, the actual performance of the machine changes with it, but the line of text on the specification sheet does not.
What the differential pressure and the flow rate change
For the same element, a different oil temperature, a different flow rate or a different differential pressure across it means a different actual velocity at the inlet. As the velocity rises, particles migrate faster across the surface of the medium and breakthrough becomes more likely; a rising differential pressure in turn shows that the medium has taken up dirt, and once it has taken up enough, the filtration accuracy may still be quoted at the original figure while the actual retention capability has already fallen.
Flow rate and model size are tied together. The flow rate of the ZYD-100 is 6000 L/h, of the ZYD-200 is 12000 L/h and of the ZYD-300 is 18000 L/h, so the latter two are at twice and three times the flow of the first. Changing the elements or the configuration within one model is in essence changing the volume of oil passing through the medium per unit of time, and as soon as that volume exceeds the design conditions of the machine the nominal accuracy no longer holds automatically.
| Parameter | Nominal value | Effect when conditions change on site |
|---|---|---|
| Filtration accuracy of the machine | ≤1 μm | No longer holds when the stage count, the sealing or the bypass state changes |
| Cleanliness | NAS 1638 class 5 | The reported value differs with a different sampling point or sampling method |
| Water content in the oil | ≤3 ppm (GB/T 7600) | Filtration itself removes no water; the water content is decided by the degassing stage |
| Fine element configuration | 150-500-5, five elements in parallel | Parallel installation improves the life, not the accuracy |
| Flow rate | 6000 L/h (ZYD-100) | At a higher flow rate, more oil passes through the medium per unit of time |
Nominal accuracy and cleanliness class are not the same thing
≤1 μm and NAS 1638 class 5 are not the same index and they describe different things. The first is the nominal retention particle size of the filter element; the second is the graded result of the particle distribution in the oil. When a machine leaves the factory both figures hold at the same time, provided it is correctly assembled, fitted with elements of the specified type, run at the specified flow rate, and the oil passing through is in an acceptable condition. If any one of these changes, re-testing the cleanliness says more than reading the specification sheet again.
- When elements are replaced, change the full set for the model rather than a single one; mixing specifications redistributes the flow between the branches.
- Before fitting a new element, check the clamping of the end cap; the symptoms of a poor seal look much like those of insufficient accuracy.
- Record the installed position and the date of each element, so that readings from different branches of the same model can be compared.
- When a drop in cleanliness is found, first confirm whether the flow rate has gone above the rated value, and only then suspect the medium.
- The oil sent for sampling and testing must come from the treated oil outlet; a sample taken in the tank carries unprocessed oil with it.
How to judge on site whether an element is performing as it should
The most direct method is to look at the differential pressure rather than at the appearance of the element. After a new element is fitted, the differential pressure should settle at a stable level and rise slowly as the run proceeds; if it is clearly higher than the previous level right after the change, check the sealing and the direction of installation first; if the same position gives the same reading on two consecutive changes, then check whether the specification of the element itself matches this machine.
The second point is to record the batches. Keep the type of each element, the replacement date, the accumulated running hours at replacement and the flow rate at the time together; after a few months it becomes clear whether this set of elements is performing consistently. Consistency is more useful as a reference than any single reading, because a single reading is affected by many site conditions.
Accuracy is a conclusion with premises. When the premises no longer hold, the figures on the sheet do not change, but what is in the oil has already changed.
