DFE . . .

The Evolution in Filter Element

Multi-Pass Development Testing
Hy-Pros Competitive Advantage

www.filterelement.com

DFE
What is DFE (Dynamic Filter Efficiency)?
All hydraulic and lube systems have a critical contamination tolerance level that is often defined by, but
not limited to, the most sensitive system component such as servo valves or high
speed journal bearings. Component manufacturers provide fluid cleanliness levels,
per ISO4406 or ISO4406:1999, required for optimum performance and predictable
life. An operating system is at risk whenever the critical contamination level is
exceeded. Contamination levels determine the individual components wear rate
(useful life) and ability to perform as intended (functionality).
System design, filter performance and maintenance practices largely determine the
contamination level in a system. Filters are expected to maintain contamination
below critical tolerance levels. Filter performance in a dynamic operating system is
variable based upon flow rate and flow density, changes in flow rate (duty cycle),
viscosity, fluid and structure borne vibration (Hz), contamination levels, ingression
rate and several other conditions. All filters are subjected to some form of system
dynamics. Hydraulic filters encounter frequent and rapid changes in flow rate
accompanied by frequency changes. Lube filters typically experience dynamic
conditions during start up and shut down. Two key characteristics of filter
performance are capture efficiency and retention efficiency. Capture efficiency can
be thought of simply as how effectively a filter captures particles while retention
efficiency is a measure of how effectively that filter retains the particles it has captured. A filter is not a black hole, and its performance must not be based solely on how efficiently it
captures particles. If not properly designed and applied, a filter can become one of the most damaging
sources of contamination in a system.
The Dynamic Filter Efficiency Test (DFE) is the evolution of hydraulic and lube filter performance testing.
The DFE test goes further than current industry standards to bridge the gap between lab and real world by
inducing dynamic duty cycles and measuring real-time performance before, during and after the cycles.
DFE testing quantifies both capture and retention efficiency in real time so that we may predict the worst
case fluid cleanliness along with average fluid cleanliness. The DFE test method was pioneered in 1998
during a joint effort between Scientific Services Inc (SSI) and Hy-Pro Filtration.

Current Filter Performance Testing Methods

Manufacturers of filter assemblies and filter elements use an industry standard test to rate filter efficiency
and dirt holding capacity of filter elements under ideal lab conditions. The test protocol is ISO standard
ISO16889 multi-pass, and was updated from ISO 4572 in 1999. The standard provides a repeatable test
method where identical filters should produce like
results when tested on various test stands. Figure 1
Contaminant
depicts the test circuit where MIL-H-5606 hydraulic
Flow meter
injected
fluid is circulated at a constant flow rate in a closed
Particle counter
loop system with the test filter and on-line particle
counters before and after the filter. Contaminated
fluid with a known quantity of contaminant is added
to the system before the upstream particle counter,
P
and at a constant rate. Small amounts of fluid are
Test filter
removed before and after the filter for particle
Reservoir
counting to calculate the filter efficiency (capture).
The capture efficiency is expressed as the Filtration
Pump
Ratio (Beta) which is the relationship between the
number of particles greater than and equal to a
Particle counter
specified size (x[c]) counted before and after the
filter.

10

www.filterelement.com

DFE
Filtration Ratio (Beta) per ISO16889:
x[c] =

quantity particles >= x[c] upstream of filter

quantity particles >= x[c] downstream of filter

Example: 7[c] = 600/4 = 150, Filtration Ratio (Beta): 7[c] = 150.

In the example, 600 particles greater than or equal to 7[c] were counted upstream of the filter and 4 were
counted downstream. This Filtration Ratio is expressed as Beta 7[c] = 150. The [c] is referred to as
sub c. The sub c is used to differentiate between multi-pass tests run per the current ISO16889
multi-pass test with new particle counter calibration per ISO11171 from ISO4572. Filtration Ratio expressed or written without the sub c refers to the
antiquated ISO4572 multi-pass test superseded by
ISO16889.
The efficiency may also be expressed as a percentage by converting the Filtration Ratio:
7[c] = 150 = (-1)/ x 100, Efficiency percentage of 7[c] = 150 = (150-1)/150 x 100 = 99.33%. The test
filter is 99.33% efficient at capturing particles 7[c] and larger.

The DFE Multi-pass Testing Method

DFE multi-pass enhances the industry standard by inducing dynamic conditions (duty cycle) and
measuring the affects of the duty cycle in real time instead of looking at normalized numbers over a time
weighted average. DFE also addresses the inherent problem of ISO16889 where fluid is added and
removed throughout the test, thus creating a small mathematical error that must be corrected in final
calculations. In addition to the capture efficiency, DFE also quantifies retention efficiency in real time. A
filter that does not properly retain previously captured contaminant can be identified. The phenomenon of
releasing captured contaminant is called unloading, and can result in temporary contamination levels that
are well above the critical contamination tolerance level of a system.
The DFE test circuit also utilizes upstream and downstream particle counters, test filter and injection point
before the upstream particle counter much like ISO16889. That is where the similarity to ISO16889 ends.
The DFE flow rate is not constant like ISO16889, but rather hydrostatically controlled so flow changes can
be made quickly while maintaining full system flow through the test filter. Particle counter sensor flows
remain constant during all particle counts and no intermediate reservoirs are used to collect the particle
counter flow before it is counted. This ensures that the fluid counted is representative of the system
contamination level. Counts are taken before, during, and after each flow change. The total number of
particle counts is determined by the duty cycle of the specific test. The efficiency results are reported in
Filtration Ratio (Beta), efficiency percentage and actual particle levels per milliliter.
The raw data is digitally tagged so filter efficiency may be reported for various combinations of flow
conditions as a time weighted average and specific ranges related to differential pressure across the filter
element. Some typical combinations include all maximum flow counts, all low flow counts and all flow
change counts (low to high or high to low). Rapid particle counting with proper timing is how DFE allows
Hy-Pro to analyze and understand both capture efficiency and retention efficiency characteristics of each
filter tested while contaminant is being introduced upstream of the filter or when there is no contaminant
being injected.

www.filterelement.com

11

DFE
The DFE Testing Method - Quantifying Contaminant Capture and Retention
Figure 2 compares the performance of two identical high efficiency glass media filter elements produced
by the same manufacturer, one of which was tested per
Filter Element
A1
A2
ISO16889 multi-pass and the other per the DFE multi-pass
method. The graph expresses the actual number of
Element Rating 7[c] > 1000 7[c] > 1000
particles 6[c] and larger counted downstream of the filter
112
112
High Flow (lpm)
element from several data points during the tests.

6
[c] particles/ml

Filter A2 was tested at a constant flow rate and maintained

a steady efficiency throughout the test. Filter A1 was cycled
between the max rated flow rate and half of rated flow
with a duty cycle consistent with that of a hydraulic
system. The downstream counts for Filter A1 varied
50
and were highest during changes from low flow to
high flow. The peaks represent counts taken during
40
flow change and the valleys represent counts
taken after each flow change. The alternating high
peaks represent counts taken during changes from
30
low flow to high flow. As the amount of contaminant
captured by Filter A1 increased, the downstream
20
counts increased most dramatically during the flow
changes from low flow to high flow. Filter element
A1, not properly designed to retain previously
10
captured contaminant during dynamic system
conditions, can become a dangerous source of
0
contamination as it captures and then releases
concentrated clouds of contaminated fluid.

Low Flow (lpm)

Contaminant
Injection Rate

56

3 mg/l

3 mg/l

Particle counts downstream of filter 6

[c]

Low ~ high flow

High ~ low flow

Figure 3 shows the particle counter raw data (top-upstream, bottom-downstream) for Element A1 before a
change from low flow to high flow and Figure 4 shows the particle counter data for Element A1 during a
change from low flow to high flow. The downstream particle count trace during the change reveals a
much higher quantity of smaller particles and larger particles that did not pass the element before the
dynamic system condition. This phenomenon can best be described as contaminant unloading. As the
filter element captures more dirt, greater amounts may be released back into the system that it is installed
to protect when the element is subjected to a dynamic flow condition and change in differential pressure
across the element. Unloading may also occur when the flow rate changes from high flow to low flow,
represented by the alternating smaller peaks in Figure 3. The filter element typically recovers shortly after
the dynamic condition, but highly contaminated clouds of fluid from contaminant unloading can cause
severe component damage and unreliable system performance.

12

www.filterelement.com

DFE
The DFE Testing Method - Quantifying Contaminant Capture and Retention
Excessive unloading in the early stage of element life may be symptomatic of an element that will eventually
fail and lose its efficiency all together (media breakdown). Filter element B (graph 9) performed true to its
rating under the ISO16889 multi-pass and achieved a beta ratio in excess of 7[c] > 1000. However, when
an identical element was tested per DFE multi-pass the beta ratio slipped well below the element rating
during dynamic conditions (graph 11). Filter media selection is often based on the beta ratio rating
published by filter manufacturers. The beta ratio is the product of the ISO16889 multi-pass test and does
not account for the dynamic duty cycle of hydraulic systems since the flow rate condition remains constant
throughout the test. A common result is a system that suffers from premature contamination related
failures, even though it is protected by filters that in theory should prevent such failures, causing reduced
uptime, unreliable equipment performance, and expensive component repair and replacement costs.
Figure 5 compares the performance of two identical Hy-Pro filter elements manufactured with G8 Dualglass
media which have been designed and developed per the DFE multi-pass test method. All Hy-Pro elements
that utilize the G8 or higher media carry the Hy-Pro DFE rating.
Hy-Pro 1

Particle counts downstream of filter 6

[c]

Hy-Pro 2

Element Rating

7[c] > 1000

7[c] > 1000

High Flow (lpm)

112

112

Low Flow (lpm)

Contaminant
Injection Rate

56

3 mg/l

3 mg/l

6
[c] particles/ml

Filter Element

Although the contaminant unloading effect is still

evident, the unloading is insignificant as filter
element Hy-Pro 1, tested per DFE, performed
true to its ISO16889 multi-pass rating of 7[c] > 1000
even during dynamic flow conditions.

Low ~ high flow

High ~ low flow

5
4
3
2
1
0

Figure 6 compares the performance of filter Element A1 and Hy-Pro 1 (DFE rated). Both elements
demonstrated excellent particle capture performance during the ISO16889 and DFE testing. The DFE rated
Hy-Pro element yielded much more stable particle counts downstream of the element and more consistent
efficiency during the dynamic flow conditions. Improving particle retention results in more predictable fluid
cleanliness levels and a system that can continually
Particle counts downstream of filter 6
[c]
operate below the critical contamination tolerance
50
level.
6
[c] particles/ml

45

40
35

Low ~ high flow

30

Filter Element

Element A1

Hy-Pro 1

Element Rating

7[c] > 1000

7[c] > 1000

High Flow (lpm)

112

112

Low Flow (lpm)

Contaminant
Injection Rate

56

56

10

3 mg/l

3 mg/l

25
20
15

www.filterelement.com

13

DFE
The DFE Multi-pass Testing Method - Cold Start Contaminant Retention
Once the element has captured enough contaminant to reach approximately 90% of the terminal P, dirty
filter indicator setting, the main flow goes to zero and the injection system is turned off for a short dwell
period. The main flow pump is turned on and rapidly achieves maximum element rated flow accompanied
by real time particle count to measure retention efficiency of the contaminant loaded element.
After the start-up simulation the system continues to perform the test duty cycle to further monitor the
retention efficiency of the filter element after a restart. The purpose of this portion of the DFE test is to
quantify how well the filter element retains the contaminant it has previously captured when subjected to a
start-up condition. The dwell before the restart may be a
Particle counts downstream of filter
function of time or a function of system temperature
8000
to simulate cold restart with an element that has
Restart unload
captured a substantial amount of contaminant.
7000
Particles per milliliter

6000
Figure 7 and the table below it show the
4[c]
performance of an element, from the same lot as
5000
filter elements A1 & A2 from figure 2, that was
6[c]
4000
subjected to the DFE restart test. During the
restart, particle counts after the filter increased
14[c]
3000
by a factor of 20 on the 6[c] channel, and the
2000
ISO codes increased by 4 on the 4[c] and 6[c]
channels. During the restart test there is no
1000
contaminant being injected so any particles
0
measured were already in the system or were
released by the element (unloading).
The result is a temporary state of
Downstream
ISO Code per
4
[c]
6
[c]
14
[c]
highly contaminated fluid that has
Element A3
particles/ml particles/ml particles/ml ISO4406:1999
resulted because the filter element
Before Restart
429
136
25
16/14/12
did not properly retain the dirt.
During Restart
6973
2802
139
20/18/14

Particle counts downstream of filter

8000

Particles per milliliter

Figure 8 and the table below it

show the performance of Hy-Pro element 3, which is
from the same lot as Hy-Pro 1 and 2 from figure 5.
The unloading is evident in the DFE rated Hy-Pro 3
element, but the affect is greatly reduced. Element
A3 (figure 7) unloaded 7 times more particles 6[c]
and larger than did Hy-Pro 3, and 35 times more
particles 14[c] and larger. The DFE rated Hy-Pro
element had much higher retention efficiency than
the filter designed and validated only to ISO16889
multi-pass.

7000

6000
5000

4[c]

Restart unload

4000

6[c]

3000

14[c]

2000

If we assume that a filter is like a black hole where

1000
all of the captured contaminant will remain trapped
0
indefinitely we are operating with a false sense of
security. If you are only discussing removal
(capture) efficiency when it
Downstream
ISO Code per
4
[c]
6
[c]
14
[c]
comes to filter elements you
Element
Hy-Pro
3
ISO4406:1999
particles/ml particles/ml particles/ml
need to be looking at particle
Before
Restart
75
10
1
13/11/7
retention efficiency as well.
During Restart

14

2994

www.filterelement.com

404

19/16/9

DFE
DFE - Comparison Between DFE and ISO 16889 Multi-Pass Test Results
Time Weighted Beta Ratio Comparison
per ISO16889 multi-pass for 5
[c] > 200
or 1000 filter element.
10000

Beta Ratio

Figure 9 shows the performance of like elements produced by

three different manufacturers that were tested per ISO 16889
multi-pass. The results were expressed as a time weighted
beta ratio. Element B had a better capture efficiency than the
Hy-Pro element in the constant flow test environment of ISO
16889. All of the elements tested were true to their Beta
Ratio of either 5[c] > 200 or 1000.

1000

100

Time Weighted Beta Ratio Comparison

per DFE multi-pass for 5
[c] > 200
or 1000 filter element.

10000

Beta Ratio

10
B

Hy-Pro

10

1000

100

10
A

Hy-Pro

Figure 10 shows the time weighted performance of the like

elements tested per DFE multi-pass. To illustrate the
performance differences between DFE and ISO16889, the two
tests were run similarly with the only difference being the DFE
test flow rate. The flow through the element was cycled up and
down the operating range to simulate a real world hydraulic
system duty cycle. The time weighted beta ratio for elements A
and B was below the rated beta ratio while elements Hy-Pro and
C performed true to rating.

Real Time Flow Beta Ratio

In figure 11 the particle counts taken during flow change have been
Comparison per DFE multi-pass for
isolated and then averaged to yield a beta ratio during transient flow.
5
[c] > 200 or 1000 filter element.
Since the DFE test has shown that filter element performance is
at its worst during flow changes isolating those sequences can
1000
help predict performance in dynamic flow systems. It is with this
graph that we see how overall filter performance can be affected
by systems with cyclic flow.
100

Element B had a beta ratio in excess of 7[c] > 2000 when tested
per ISO16889 (figure 9). However, figure 11 shows the average
beta ratio of Element B during variable flow to be less than
7[c] > 100. The Hy-Pro element beta ratio was in excess of
7[c] > 800 and was the only one with a beta ratio greater than
100. The Hy-Pro performance in figure 11 illustrates why Hy-Pro
is committed to the DFE test method for design and development.

Beta Ratio

11

10

1
A

Hy-Pro

Relying solely on ISO16889 to predict how filter elements will perform in systems with dynamic flow
conditions means that we are making decisions on filter performance without all of the available information.
The current industry standard test for hydraulic and lube filter performance (ISO 16889) is a good tool for
predicting performance of off-line filters and circulating systems, but does not accurately represent the
stress of a hydraulic circuit with dynamic flow conditions or a lube system cold start condition. The first step
to fixing a problem is acknowledging that a problem actually exists, and without DFE testing it is difficult to
truly predict actual filter performance in a dynamic system.

www.filterelement.com

15

DFE
Understanding ISO Codes - The ISO cleanliness code (per ISO4406-1999) is used to quantify
particulate contamination levels per milliliter of fluid at 3 sizes 4[c], 6[c] and 14[c]. The ISO code is
expressed in 3 numbers (example: 19/17/14). Each number represents a contaminant level code for the
correlating particle size. The code includes all particles of the specified size and larger. It is important to
note that each time a code increases the quantity range of particles is doubling.

ISO 4406:1999 Code Chart

Range
Code
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6

Particles per milliliter

More than
Up to/including
80000
160000
40000
80000
20000
40000
10000
20000
5000
10000
2500
5000
1300
2500
640
1300
320
640
160
320
80
160
40
80
20
40
10
20
5
10
2.5
5
1.3
2.5
0.64
1.3
0.32
0.64

Particles
Particle
ISO 4406
per
milSize
Code range
liliter
151773 80000~160000
4
m[c]
38363
20000~40000
6
m[c]
8229
10m[c]
3339
2500~5000
14
m[c]
1048
21m[c]
112
38m[c]

Particle
Size
4
m[c]
6
m[c]
10m[c]
14
m[c]
21m[c]
38m[c]

Particles
per
milliliter
492
149
41
15
5
1

ISO
Code
24
22
19

ISO 4406
Code range

ISO
Code

320 ~ 640
80 ~ 160

16
14

10 ~ 20

11

Succeed with a Total Systems Cleanliness Approach

Developing a Total System Cleanliness approach to control contamination and care for fluids from arrival to
disposal will ultimately result in more reliable plant operation and save money. Several steps to achieve Total
Systems Cleanliness include: evaluate and survey all hydraulic and lubrication systems, establish an oil
analysis program and schedule, insist on specific fluid cleanliness levels for all new fluids, establish a baseline
and target fluid cleanliness for each system, filter all new fluids upon arrival and during transfer, seal all
reservoirs and bulk tanks, install high quality particulate and desiccant breathers, enhance air and liquid
filtration on existing systems wherever suitable, use portable or permanent off-line filtration to enhance existing
filtration, improve bulk oil storage and handling during transfer, remove water, and make a commitment to fluid
cleanliness.
The visible cost of proper contamination control and total systems cleanliness is less
than 3% of the total cost of contamination when not kept under control. Keep your
head above the surface and avoid the resource draining costs associated with fluid
contamination issues including:

Downtime and lost production

Component repair/replacement
Reduced useful fluid life
Wasted materials and supplies ($)

16

Root cause analysis meetings

Maintenance labor costs
Unreliable machine performance
Wasted time and energy ($)

www.filterelement.com