Even a small amount of stress can have a large effect. Driving in a massive traffic jam may cause you to act differently, perhaps becoming hostile to other drivers, or even bringing on a headache. Speaking to a large group of people may cause a person to repeatedly say, “uhhh,” while his or her palms sweat profusely. It’s safe to say that we often act differently under stress. Similarly, components in a fluid power system running at full power in a dirty factory may behave in ways quite different from ones sitting on a simple test bench.

In a stressful operating environment, filter integrity and optimum performance depend on other significant parameters:

Filters, critical for contamination control in hydraulic systems, can certainly exhibit different qualities when placed under stress. To ensure that a filter is correct for an application, its performance must be qualified using appropriate and rigorous testing protocols. Protocols should employ parameters to measure performance. Unfortunately, however, filter comparisons and selection are based primarily on filtration ratio and contaminant-holding capacity as determined by a single multi-pass test (ISO 16889).
● strength and stability of the filtration medium with respect to contaminant loading under the operating fluid and system conditions
● ability to withstand flow and pressure surges, and
● conditions induced by cold start-ups.

Because the scope of multi-pass testing is limited, it does not provide sufficient information to assess filter performance in relation to these criteria. As a result, weaknesses and deficiencies in performance may be hidden or overlooked.

The multi-pass test

During the multi-pass test, a controlled rate of contaminant is injected into the filter test system, while online particle counts are continuously taken from fluid samples both upstream and downstream of the test filter.

The contaminant specified is ISO Medium Test Dust, an irregularly shaped, silica-based material, with a particle size ranging from sub-micrometer to about 80 μm. The fluid is MIL-H-5606 aircraft hydraulic fluid, with an anti-static additive to control fluid conductivity. Contaminant is injected at a high rate, causing the filter to become plugged, after which the test terminates (typically in 0.5 to 2 hrs). The filter is tested at its rated flow, which is maintained constant throughout the test.

The measure of the filter’s ability to remove contaminant is its filtration ratio, also called the beta (b) ratio. This is defined as the number of particles of a specified size upstream of the filter relative to the number of particles of the same size downstream of the filter. The multi-pass test also rates a filter on its contaminant- holding capacity (how much it can hold prior to reaching terminal differential pressure).

Limitations — Although multipass has improved on earlier nominal methods, its deficiencies preclude it from truly representing a filter’s performance during service in the field, especially under variable flow and pressure operating conditions. The following differences between laboratory and actual service conditions explain why performance during the multi-pass test might surpass actual performance during operation:
● The multi-pass test is conducted under steady-state conditions. Practical fluid system operation is rarely that controlled or stable, and variable flow is integral to system operation.
● Instead of the allotted test time being equivalent to the duration of a filter’s actual service life — say, three to 12 months — the test is conducted in only a few hours. Unrepresentative contaminant concentrations (1,000- to 10,000-fold higher than actual) are used for the test.
● ISO MTD is used in the standard industrial multi-pass test. It has a well-controlled particle size distribution, but it is not equivalent to the contaminant in the actual operating system.
● The test fluid, with its low 15-cSt viscosity, is not representative of the fluid used in the operation of most industrial fluid power systems. In addition, an anti-static additive is not used during actual service.
● The filter element is not exposed to cold start-up conditions or vibration during testing. Both of these can be expected in many actual applications.

Realistic testing conditions — The only accurate test of a filter’s capability is that performed during actual service. Since this is impractical, it is important that conditions in the laboratory are as equivalent as possible to those present during actual operation. Filter testing protocols must be sufficiently detailed to expose deficiencies in housing and element design.

Recently, ISO acknowledged that filter selection should not be based upon data from a single multi-pass test, and it introduced a standard for filter performance, ISO 11170. ISO developed a range of laboratory tests in which the filter element is exposed to severe conditions, such as extreme temperatures and cyclic flow. The filter is first subjected to a general compatibility test using cold and heat soaks at minimum and maximum operating temperatures.

After passing the compatibility test, three elements are subjected to a full range of ISO tests. A complete filter element specification should include performance ratings for:
●flow fatigue resistance, with reference to the number of cycles and to peak-to-peak differential pressure (∆P), and
●maximum clean pressure drop at system flow rate and fluid viscosity.

Impact of unsteady flow — Most early studies addressing the impact of cyclic flow on filter performance during the multi-pass test were conducted at Oklahoma State University’s Fluid Power Research Center. When the multi-pass test was used for these studies, the flow was cycled uniformly in a roughly sinusoidal manner, from zero to full flow, at varying frequencies. The results indicated that the filtration performance of all filters tested decreased as a function of increasing cycle rate.

A series of tests used flow surges with a square waveform to demonstrate that the filtration ratio decreases when a flow surge is applied. In most cases, the filtration ratios were inversely related to the magnitude of the flow surge. It was theorized that desorption of previously captured contaminant was the primary cause for the increased downstream particle count.

A survey of mobile equipment manufacturers that used suction- and return-line filters revealed that nearly all installations were subject to variable flow. The average frequency of flow surges was 0.093 Hz, or 10.8 sec/cycle. The surge magnitude, as a multiplier on minimum flow, varied from 1.33 to 8.13, with an average of 2.5.

The flow fluctuations encountered by operating hydraulic filters impose an energy pulse on the particles captured by the filter medium. Depending on the filter design and capture mechanisms, this energy can displace a particle from its capture site and allow it to pass downstream. This is often referred to as contaminant desorption or unloading.

A filter that has a uniform and fixed pore size and depends on direct interception or mechanical blockage as the primary capture mechanism tends to hold on to particles. A non-uniform filter medium, which is generally thicker and depends on adsorption (particles sticking to the fibers) for particle capture, has the tendency to desorb particles in the presence of variable flow and its resultant forces.

Cyclic Stabilization Test

A laboratory testing method has been developed to resolve many of the deficiencies of the Multi-pass test. The Cyclic Stabilization Test (CST) examines cyclic flow performance and the effects of contaminant loading on the retention and desorption of the filter.

During the development process, numerous experiments were conducted to set parameters and determine the optimum times for injection, particle counting, and contamination reporting. The final version of the CST is conducted in a similar manner to the multi-pass test, using virtually the same equipment. The CST requires the addition of a bypass line and a flow cycling valve to continuously cycle the flow through the filter. Filtration ratios are measured as contaminant is introduced to load the filtration medium and increase the differential pressure. At two points in the life of the filter, the cleanup and stabilization (drawdown) characteristics are measured by stopping the contaminant injection for typically a 30-min period while maintaining cyclic flow conditions. This 30-min period is called the drawdown or stabilization period.

A great deal of information can be gathered from running the CST. The following measurements can be obtained, in this order, during the testing process:
● initial cyclic flow filtration ratio,
● cyclic flow filtration ratio (throughout the remainder of the test) and average filtration ratio under cyclic conditions,
● cleanup and stabilization when the filter is partially loaded with contaminant (2.5 % ∆P increase),
● cleanup and stabilization at 80% ∆P increase (conducted while the filter is almost fully loaded),
● dirt-holding capacity, under cyclic conditions.

The test stand is basically the same as in the multi-pass, except that an electrically controlled valve allows flow to be cycled from 25 to 100% of the filter’s rated flow. The test cycle rate is 0.1 Hz, or 6 cycles/min, with a rise time and fall time from 0.2 to 0.5 sec. A survey of original equipment manufacturers concluded that this rate is widely applicable. Nevertheless, if desired, the test may be conducted using any cyclic condition.

Cyclic flow filter testing

The SAE A-6 committee on aerospace hydraulics standards is developing a test method for measuring the performance of hydraulic filters under cyclic flow conditions. Its draft procedure, SAE 4205, is entitled Aerospace Fluid Power — Hydraulic Filter Elements — Method for Evaluating Dynamic Efficiency with Cyclic Flow. It proposes a test that expands on the standard multi-pass by introducing cyclic flow and stabilization drawdown periods with no contaminant ingression. It is equivalent to the CST, except that it uses ISO Fine Test Dust (ISO FTD) rather than ISO MTD, which is used in the industrial procedures. The aerospace procedure is currently undergoing ballot at the SAE panel level.

NFPA/ISO projects — The NFPA has also proposed an alternative test procedure to the multi-pass test. This draft procedure has been proposed to ISO and has been accepted as a new project, ISO WD 23699, under the ISO/TC131 committee. It is very similar to the SAE ARP 4205 draft, including cyclic flow at a rate of 0.1 Hz, and also to the CST proposed below. It does not, however, currently include drawdown or stabilization periods.

ISO WD 23699 must undergo three successful ballots at the international level before becoming an ISO standard, a process that typically takes several years.

Comparison of test procedures

Performance comparisons of filtration ratios resulting from the different test procedures (standard multi-pass test, CST, and CSTV) were conducted on three different sets of filters, A, B, and C. (CSTV is the CST test with vibration added to the test filter.) These results demonstrate that the multi-pass test method qualifies the upper bound performance for a filter, whereas the CST and the CSTV tests qualify the lower bound performance for that filter.

Recommendations — The testing detailed in this article has demonstrated that the CST is a repeatable and reproducible method for evaluating the effects of cyclic flow and vibration on the performance of hydraulic filter elements. Using the CST, the dynamic efficiency and contamination control ability of the filter can be measured in a manner that more closely represents the operation of most hydraulic systems.

The CST method has the following advantages:
● It can be performed on a slightly modified multi-pass stand, of which there are over 100 in the world.
● It is repeatable and reproducible between labs.
● It indicates the average contamination level maintained in the actual cyclic flow system. This average level could be related to component wear rates.

The test results presented reveal the following:
● All filters tested have a lower performance with cyclic flow than is indicated using the standard steady-flow multi-pass test. Therefore, the CST represents the lower bound on performance.
● Both the magnitude of the flow cycles and cyclic frequency have a major effect on the stabilized downstream particle count. Standard test conditions are specified for the CST method.
● For the filters tested, vibration had little effect on performance. However, this is most likely not true for all filters.

Users of hydraulic filters should use the CST to better understand the expected performance of hydraulic filters and to specify the desired performance. I also suggest that the standards organizations continue their efforts to adopt a standard cyclic flow filter test method.