K&N / Pure Power put on a dyno, on a Subaru

[flstffxe]

Sorry for the long post but there is a lot of info and I tried to organize it as best as possible.

Back story, acquaintance went to get a baseline on his current setup before doing future mods. I had a K&N SS-1008/Pure Power 8415 that was not in use. Since he was going this was a convenient chance to see if a free flowing stainless steel micronic screen filter has any affect on HP/TQ. Remember it takes HP/TQ to run a oil pump.

The oil pump on this engine is a 12mm JDM pump with one shim on the oil pressure relief valve. Factory 12mm pump flow and relief valve specs are below:
600rpm = 1.717 gallons per minute @ 14psi
6,000rpm = 15.586 gallons per minute @ 85psi(capped at relief valve pressure)
Relief valve pressure 85psi(IIRC one shim should raise this ~5-8psi)

Keep in mind this engine was spinning to 7,000rpm.

He had what I would guess to be a 20-30 mile drive there, parked for 15 minutes, then put on the dyno for his 3 baseline pulls.

Pulled it off the dyno(backed it up 25") swapped out the 152208AA12A(OEM blue) for the primed K&N/PurePower set it back up on the dyno and did 3 more pulls. Down time for the filter swap was about 20 minutes. So the pulls are about as consistent as any one could make them. The heat soak for the engine should have been virtually identical between the first set of 3 pulls and the second set of 3 pulls.......

Results are not at all what we expected and well, interesting to say the least.

Pertinent mod list in regards to the powertrain:

• Subaru EJ22t
• 9:1 Compression CP pistons
• Eagle rods
• ’96 Impreza crank (Phase 1)
• ACL bearings
• FHI 12MM Oil Pump (1 extra shim)
• 2002 WRX heads
• Port and Polished
• Combustion chamber matched to block (50CC chamber)
• GSC Stage 1 Cams (266I/266E)
• Manley springs and retainers
• Unorthodox Racing underdriven crank pulley
• Agency power parallel fuel rail system
• Deatchwerks 1000CC/min injectors
• Walboro 255LPH fuel pump
• SPT Intake & Filter
• ’04 STi TMIC
• Forced Performance HTA68 Turbo
• Custom unequal length exhaust manifold
• Invidia catless uppipe
• Helix catless downpipe
• SPT catless midpipe
• STi Genome axle back
• OpenECU (Nuke) Tune
• Running BP 93 octane gasoline
• Rotella T 5w40, changed a week prior

Driveline Stuff:

• Exedy Hyper Single flywheel, clutch, pressure plate
• ’04 STI 6 Speed & R180
• 17” Prodrive P7s
• 225/45/17 Hankook V12s

We did not look at or know the results prior to telling the dyno operator what runs we wanted on each print out and him handing them to us. Logically we chose to follow the obvious:

- Put the 1st set of 3 runs(OEM filter) on one graph
- Put the 2nd set of 3 runs(K&N filter) on one graph
- Put the last run of each set on one graph as the comparison.

The logic in using the last run of each set is that each one had 2 previous runs before it in quick succession thus every thing should be equally heat soaked.

As I am not one to pull any punches the graphs below are posted in the above order and there is nothing to hide. Links to full res/size images are below embedded pics.


1st set of 3 runs(OEM filter)

http://img59.imageshack.us/img59/4716/dynoruns13stockfilter65.jpg
RUN 1 HP 285 TQ 281
RUN 2 HP 263 TQ 258
RUN 3 HP 260 TQ 251


2nd set of 3 runs(K&N filter)

http://img266.imageshack.us/img266/2634/dynoruns46knfilter65201.jpg
RUN 1 HP 281 TQ 274
RUN 2 HP 283 TQ 276
RUN 3 HP 278 TQ 266

last run of each set on one graph as the comparison

http://img266.imageshack.us/img266/2634/dynoruns46knfilter65201.jpg
1st set, RUN 3 OEM Honeywell/Subaru 152208AA12A Filter HP 260 TQ 251
2nd set, RUN 3 K&N SS-1008 (AKA Pure Power 8415) Filter HP 278 TQ 266

The 1st set of 3 baseline runs lose 25HP and 30Ft-Lbs from the first run to the third run. One would normally attribute this to intercooler heat soak.

The 2nd set of 3 K&N filter runs lose 3HP and 8FT-Lbs from first run to third run. What happened to the intercooler heat soak theory we had?

The difference from the last run with the OEM *12A filter(HP 260 TQ 251) to the last run with the K&N SS-1008 filter( HP 278 TQ 266) is 18HP and 15Ft-Lbs.

1st set, RUN 1 OEM Honeywell/Subaru 152208AA12A Filter HP 285 TQ 281
2nd set, RUN 1 K&N SS-1008 (AKA Pure Power 8415) Filter HP 281 TQ 274

There were no other changes between the two sets of pulls other then the oil filter being changed. Both runs had 3 blowers pointing at the radiator and one directly strapped to the intercooler in the same spot for each set.

I am not going to jump to conclusions on this one. The only one I have is that some thing happened. Just not sure what.

My only wish is that we would have gotten oil temps also.

So what do you make of the results?

 

[bepperb]

The standard deviation is so high it's tough to do anything with those numbers (correlation would be strongest with the run number, not with the filter change). Heat soak certainly seems to be an issue, that's not something that's going to be impacted by a filter. It would be great to have some intake air temp readings the next go around, and some spark advance numbers as well (to see if the computer is retarding ignition due to heat causing precombustion). BP 93 should be good but who knows.

Sorry, I have as many questions as you do... but no faith in the filter having done anything.

 

[flstffxe]

Don't get me wrong I would have liked to log every thing, however we were only really there for him to get hos baseline #s before he went any further with the car. No tuning, so we didn't see a need to log. Kind of wish we had so we could see what's with the huge variance in the first set.


There are other places to make power then hoping for it from a filter. Doing a set with the K&N filter was just a convenient opportunity to put some numbers behind some of these manufactures "potential" claims.

There wont likely be a next round of filter VS filter on a dyno, at lest I don't plan on it or see a need for it.

IMHO the first run from each set, disproves the possible gains theory.



1st set, RUN 1 OEM Honeywell/Subaru 152208AA12A Filter HP 285 TQ 281
2nd set, RUN 1 K&N SS-1008 (AKA Pure Power 8415) Filter HP 281 TQ 274

Those numbers are well with in what I would expect for a typical run variation.

 

[blmlozz]

Originally Posted By: flstffxe


Those numbers are well with in what I would expect for a typical run variation.

+1, I have never seen any car duplicate exactly the same dyno graph twice. Even on the same dyno back to back.

Do we even have flow-rate data on the K&N compared to any other filter of the same model?

 

[flstffxe]

Originally Posted By: blmlozz
Originally Posted By: flstffxe


Those numbers are well with in what I would expect for a typical run variation.

+1, I have never seen any car duplicate exactly the same dyno graph twice. Even on the same dyno back to back.

Do we even have flow-rate data on the K&N compared to any other filter of the same model?


All the other aftermarket filters which list flow rate are in the 7-12 gallons per minute range. Subaru has/will not release the specs for the current/past OEM filters in regards to flow.

The K&N(SS-1008)/Pure Power(8415) is listed at 22gpm.

 

[SteveSRT8]

Originally Posted By: blmlozz
Originally Posted By: flstffxe


Those numbers are well with in what I would expect for a typical run variation.

+1, I have never seen any car duplicate exactly the same dyno graph twice. Even on the same dyno back to back.

Do we even have flow-rate data on the K&N compared to any other filter of the same model?


X a dozen! I have hundreds of dyno runs across years and would absolutely agree.

Add that the numbers are highly suspect as well.

 

[ZeeOSix]

Originally Posted By: bepperb
The standard deviation is so high it's tough to do anything with those numbers (correlation would be strongest with the run number, not with the filter change).


Yes, seems to be too much variation in back-to-back pulls. In a simple sense, you could average the 3 pulls with each filter to get a feel for any impact of using the stainless mesh filter.

OEM Filter, 3 pulls avg = 269.3 HP

Stainless Mesh Filter, 3 pulls avg = 280.7 HP

Delta = +11.4 HP with SM filter.


Originally Posted By: flstffxe

The oil pump on this engine is a 12mm JDM pump with one shim on the oil pressure relief valve. Factory 12mm pump flow and relief valve specs are below:
600rpm = 1.717 gallons per minute @ 14psi
6,000rpm = 15.586 gallons per minute @ 85psi (capped at relief valve pressure)
Relief valve pressure 85psi (IIRC one shim should raise this ~5-8psi)

Keep in mind this engine was spinning to 7,000rpm.


The stainless mesh filter will certainly have less PISD across the filter when 15+ GPM is flowing through it as compared to the OEM filter. How much is unknown unless you actually had a PSID vs GPM flow curve of the OEM filter. The PureOne that Purolator bench tested for SuperBusa showed it had around 7 PSID at 15 GPM with hot oil.

http://www.bobistheoilguy.com/forums/ubb...451#Post1619451

But if the stainless mesh filter only had say 2 PSID at 15 GMP, then there is definitely some HP being saved that would normally be used to pump the oil through more restriction. I could believe a few HP savings with a near zero restrictive oil filter, but not sure if 11 HP is accurate ... could be though, as it takes more power to pump oil than most realize.

The level of HP difference could actually be calculated pretty easily if you knew what the actual oil pump output pressure was in each case. Most oil pressure gauges however show the oil pressure between the oil filter and the engine, and that shouldn't show a change between the two filters unless the oil pump was in pressure relief mode. With the stainless mesh filter, I doubt the oil pump hit pressure relief ... but it could have with the OEM filter. Was there an oil pressure gauge, and was it recorded during the dyno runs?

 

[boxcartommie22]

the k&p ss filter flows alot more(57gpm) then k&n ss 20gpm

 

[ZeeOSix]

Originally Posted By: boxcartommie22
the k&p ss filter flows alot more(57gpm) then k&n ss 20gpm


Are those flow rate specs at some defined viscosity and PSID across the filter? They don't mean much unless there is some kind of qualifier associated with the flow number. Kind of like saying a filter is "99%" efficient without saying at what micron size.

 

[flstffxe]

They are actually fairly close. The K&N/Pure Power are greatly under rated. They should be within about 10 gallons of each other. K&P = 35u
K&N = 22u

From work I don't recall the difference in flow between a 10u screen and a 100u screen being much at all. A few(single digits) gallons per minute with a 20wt silicone fluid.

ZeeOSix, I agree. As I mentioned we didn't care about logging as we were not there for tuning. So no oil pressure data.


A real test would in my opinion actually require setting a Subaru pump up on a test stand driven via an electric motor that allowed you to set and monitor the load, flow, pressure....

 

[ZeeOSix]

Originally Posted By: flstffxe
A real test would in my opinion actually require setting a Subaru pump up on a test stand driven via an electric motor that allowed you to set and monitor the load, flow, pressure....


Could do a rough calculation to find the HP difference of pushing 15 GPM of oil at X viscosity with a supply pressure of 85 PSI vs a supply pressure of 80 PSI. The 5 PSI difference in required supply (pump) pressure to move the 15 GPM would represent the difference in filter PSID between the two filters.

 

[ZeeOSix]

http://en.wikipedia.org/wiki/Horsepower

"Hydraulic horsepower is equivalent to mechanical horsepower. The formula given above is for conversion to mechanical horsepower from the factors acting on a hydraulic system."

Hydraulic horsepower = flow rate (US gal/min) × pressure (psi) × 7/12,000

Using 15 GPM and 85 PSI: HP = 0.74

Using 15 GPM and 80 PSI: HP = 0.70

Humm ... doesn't seem to make much difference, and surprisingly from the equation it doesn't take that much HP to move 15 GPM at 85 PSI.

Another calculation source:
http://www.calcunation.com/calculators/machinery/fluid power/hydraulic-horsepower.php

 

[boxcartommie22]

upto 70wt oil 35u absolute.