What's new
Van's Air Force

Don't miss anything! Register now for full access to the definitive RV support community.

Crankcase Vent, oil on belly

n707sm

Member
Crankcase Vent, oil on belly & speed test CC vent in/out of cowl

My RV-7a firewall forward sports a Barrett I-O 360 (with ~80 hours), a horizontal intake, a Sam James Cowl with a Sam James Plenum, Vetterman 4-into-2 cross-over exhaust, and Robert Paisley's EFII (dual ecu's). I'm very happy with this setup, and the engine runs strong. However, the belly is oil-coated at the rate of about 1 qt every 5-6 hours, and that's with only starting at 6.5 qts per change/fill. I reached out to Allen Barrett a couple of weeks back, and his suggestion was that the oil is being "pulled" or expelled out somehow, and to look into that first.

After several hours of searching/reading the many posts on VAF about oil on the belly, the one post that seemed more likely to be a match for my situation was this excellent post by Tom Martin a number of years back.

To determine if I have the same issue, I enlisted the help of my good friend Lee (the crazy Hawaiian dude who's also a great A&P as well), where we set up three measuring locations using his magnahelic: Upper cowl (inside plenum), lower cowl (at CC vent) and crankcase (by way of oil filler tube). After checking to make sure all test locations were secure, and could handle full power runs, I took some measurements. All measurements were at 7500 MSL, 2600 RPM with 23" MP, using a 2-axis auto-pilot to hold things steady while I swapped out the tubes on the magnahelic. A few tests were done during climb, but the data was not significantly different, so I'm only posting the level flight measurements. Also, the process of making a change, and re-testing spanned a couple of days, so the OAT's were not identical from test-to-test, but all speeds are measured in Knots True from my trusty Dynon (with OAT probe). Prior to taking these measurements, I cleaned the airplane which includes a spotless belly.

This first picture shows the location and configuration of the CC vent (along with a diffuser on the measurement line) before any changes. Basically Van's stock setup (but without the recommended slant-cut on the end of the tube -- I missed that subtle detail..).

image_zpsar1mthvu.jpg


Test Case 1: No changes, original setup
Location/HG Measurement DATA:
Crankcase: OAT 65F,171KTS,7500MSL,2600RPM,23MP,2.7"WC
Bottom: OAT 65F,171KTS,7500MSL,2600RPM,23MP,2.5"WC
Top: OAT 65F,171KTS,7500MSL,2600RPM,23MP,10+"WC (pegged the 0-10 HG Magnahelic)

From the above data I was encouraged that my scenario is matching what Tom Martin saw (however, my pressures were lower, but still relative relationship-wise), where the crankcase pressure was slightly higher than the pressure measured low in the cowl at the CC vent.

This next picture is where I chose to take a baby-step, and change the cut on the end of the CC vent tube to match Van's plans. This is the only change made between these tests.

image_zpsdobowhh6.jpg


Test Case 2: Slant cut on CC vent tube
Location/HG Measurement DATA:
Crankcase: OAT 72F,171KTS,7500MSL,2600RPM,23MP,2.6"WC
Bottom: OAT 72F,171KTS,7500MSL,2600RPM,23MP,2.7"WC
Top: OAT 72F,171KTS,7500MSL,2600RPM,23MP,10+"WC (pegged the 0-10 HG Magnahelic)

Interesting numbers here, and I suppose I could have stopped with my changes to see if the oil on the belly would be reduced simply by the slant cut on the tube end. I decided to keep testing, because I really wanted to see what would happen by moving the CC vent outside of the cowling -- which these next two pictures show....

image_zpsgh0cjo9z.jpg


image_zpsu5ljiy6j.jpg


Now, I was quite surprised by these numbers:

Test Case 3: CC vent tube outside cowling
Location/HG Measurement DATA:
Crankcase: OAT 72F,175KTS,7500MSL,2600RPM,23MP,-1.3"WC
Bottom: OAT 72F,175KTS,7500MSL,2600RPM,23MP,-.5"WC
Top: OAT 72F,175KTS,7500MSL,2600RPM,23MP,10+"WC (pegged the 0-10 HG Magnahelic

Wow! A 4KT increase in speed! (it actually was bouncing between 175-176KTS). I wasn't going for that, but needless to say, I was very pleased. Immediately after this test I inspected the belly for oil, and also removed the test tube from the Oil filler tube, and double-checked the oil level -- yep, same as where it started. I've since put 2 more hours on the aircraft and I am checking the oil level before/after each flight, as well as inspecting the belly. So far, still clean on the belly. In addition, I called Mr. Allen Barrett and shared these findings with him, and to get his opinion if there is any risk to continuing the testing with a negative pressure on the CC vent. He saw no reason to not continue, but was cautious to keep checking the oil levels before and after each flight. So, I will keep flying/checking, and will report back!

Michael
 
Last edited:
So, did a slight vacume on the crankcase help to seal the piston rings and make more HP?
 
@Bret,

I'm not an engineering-type, nor do I have any training or experience in that area -- but, from my research, and reading the various discussions on this forum, the general thinking appears to be a) Lower windage in the CC with negative pressures, and b) ring flutter at the bottom is reduced/eliminated.

My theory is that with the rings tight against the cylinder wall at the bottom, you get a complete compression stroke without having to loose valuable stroke distance (at the bottom, on the way back up) getting the rings seated again.

I would love to hear from any experts that wouldn't mind chiming in...

Michael
 
This makes total sense now, during combustion, the gases go down the outside of the piston crown, to the gap on top of first piston ring land and and top of ring, to the inside of piston ring and outside of inner wall of piston ring land and seals on the bottom of the top ring and bottom surface of top piston ring land and pushes the piston ring out to the cylinder wall, if there is any pressure on the bottom surface of the top ring it COULD unseat the ring from the wall, causing more blowby ?........I think ?.......someone check my insanity please....
 
A few accuracy notes....

If you did not connect the aircraft static system to the static port of the gauge, the gauge pressure is the difference between cockpit pressure and probe pressure. The actual pressures at the probe locations may thus be higher or lower than indicated. As an aside, cabin air vent open or closed can also change the indication.

If you did connect the static port of the gauge to aircraft static, you could easily affect your airspeed indication; the probe hanging behind the cowl exit is probably in a lower pressure region than the fuselage static ports. That's why it's necessary to fly a NTPS GPS three-leg if valid airspeed is important to the test...the recorded airspeed data is independent of the aircraft's pitot static system.

There is a HP gain with negative case pressure, but I'd be surprised if it was enough to pick up 4 knots, in particular if it was only 1.3" H2O negative.

A valid test is repeatable.
 
Last edited:
@Dan,

Very good points. I was using cabin pressure, with the same configuration of cabin vents (between tests), but I do understand what you mean about the accuracy is in question. Between test case #2 and #3 there was roughly 3 hours time on the same day with relatively the same conditions. Both tests were in the afternoon, and both tests flown over the same area for time and distance. Test case #1 was the previous day in slightly cooler conditions.

Whether I will see a speed gain (or not) over time is secondary to my primary objective of keeping the oil in the crankcase and off of the belly, but I should be able to tell over the next 10 hours or so if this plane is noticeably faster than before.

Very much appreciate the feedback!
Michael
 
I am more interesting in hearing how you lose less oil through the breather with a lower crankcase pressure. Seems backwards to me.
 
I know! ...it's a head-scratcher for sure ;-) I will report back once I have more time with this new configuration.
 
Whether I will see a speed gain (or not) over time is secondary to my primary objective of keeping the oil in the crankcase and off of the belly, but I should be able to tell over the next 10 hours or so if this plane is noticeably faster than before.

Like I said, a valid test is repeatable. Re the speed increase, just move the breather inside and outside the exit, and fly some 3-leg laps.

http://www.ntps.edu/images/stories/documents/gps-pec-method.doc

http://www.ntps.edu/images/stories/documents/gps-pec.XLS
 
Thanks Dan for the links to the reference doc, and spreadsheet. I will see about doing the 3-leg testing with vent both in and outside of the cowling. The weather is forecast to get rainy here later this week and over the weekend, so it may be a bit before I find decent conditions (with a stable/consistent air mass) over (or near to) Tahoe any time this week.
 
Engine Mount

May I suggest the removal of the zip tie from the motor mount and the use of an adel clamp instead? Sure looks clean in there.
Don
 
Thanks Dan for the links to the reference doc, and spreadsheet. I will see about doing the 3-leg testing with vent both in and outside of the cowling. The weather is forecast to get rainy here later this week and over the weekend, so it may be a bit before I find decent conditions (with a stable/consistent air mass) over (or near to) Tahoe any time this week.

Yer' welcome.

It's really easy with a 2 axis autopilot and calm air. Put it in heading bug mode with alt hold engaged, let everything settle down, record track and GPS groundspeed, twist the bug about 120 and repeat, do it a third time, and make the spreadsheet entries when you get home.

Be sure to record track, not heading.
 
May I suggest the removal of the zip tie from the motor mount and the use of an adel clamp instead? Sure looks clean in there.
Don

That's a good suggestion, and with the recent change to move the vent outside, it now has adel clamps instead. The photos of course still show the *temporary* zip ties to hold the measurement line and diffuser, and those are now gone as well.

I hope it stays clean! ...thanks :D
 
Yer' welcome.

It's really easy with a 2 axis autopilot and calm air. Put it in heading bug mode with alt hold engaged, let everything settle down, record track and GPS groundspeed, twist the bug about 120 and repeat, do it a third time, and make the spreadsheet entries when you get home.

Be sure to record track, not heading.


Gotcha, will do! :)
 
... The weather is forecast to get rainy here later this week and over the weekend, so it may be a bit before I find decent conditions (with a stable/consistent air mass) over (or near to) Tahoe any time this week.

I'm hoping the weather takes a turn for the better- I'm planning a trip to RNO this weekend in the Rocket...

Anyway - does your vent tube just dump out the side with the standard Lyc fitting, or is there any attempt on your part to get some additional height before it turns and heads out the cowl? That additional vertical height (and an standard automotive breather) made a huge difference on my Rocket. I've been coast to coast twice and there is virtually zero loss out the breather any more. It just stopped.
 
I'm hoping the weather takes a turn for the better- I'm planning a trip to RNO this weekend in the Rocket...

Anyway - does your vent tube just dump out the side with the standard Lyc fitting, or is there any attempt on your part to get some additional height before it turns and heads out the cowl? That additional vertical height (and an standard automotive breather) made a huge difference on my Rocket. I've been coast to coast twice and there is virtually zero loss out the breather any more. It just stopped.

I'm with you on something good from the weather gods!

The vent just comes out standard in a horizontal plane. I like the suggestion to take it vertical some first, and will go that route as the next change if needed. I might go that route anyway, but will not make any other changes for the next 10 hours or so.

...(and a standard automotive breather)

Are you referring to some sort of PCV valve?
 
I'm with you on something good from the weather gods!

The vent just comes out standard in a horizontal plane. I like the suggestion to take it vertical some first, and will go that route as the next change if needed. I might go that route anyway, but will not make any other changes for the next 10 hours or so.



Are you referring to some sort of PCV valve?


Nope. Just a breather. It's discussed here: http://www.vansairforce.com/community/showpost.php?p=973619&postcount=10
 
potential racing mod

looks like a potential racing mod. more delta P across the piston = more power. for racing, don't care about oil on the belly or cold weather freezing water in the tube during winter ops.
 
Last edited:
What is the ICP in a Lycoming cylinder? I can't find data online.

Anyway, the point is that the ICP has to be hundreds of PSI. I can't imagine that reducing the crankcase pressure by 1.4 PSI would result in a substantial power increase, better ring seating, or any improvement that is measurable without extremely sensitive test equipment.
 
Anyway, the point is that the ICP has to be hundreds of PSI. I can't imagine that reducing the crankcase pressure by 1.4 PSI would result in a substantial power increase, better ring seating, or any improvement that is measurable without extremely sensitive test equipment.

HP increase is due to reduction of windage and pumping losses.

I'm guessing you meant 0.14 psi (1.3"+2.6" H2O)...pesky decimal in the wrong place.

For comparison, exhaust evacuator users are pulling about 10 times that much case vacuum. I was in a dyno room a few weeks ago with an engine builder who has measured a HP increase.
 
HP increase is due to reduction of windage and pumping losses.

I'm guessing you meant 0.14 psi (1.3"+2.6" H2O)...pesky decimal in the wrong place.

For comparison, exhaust evacuator users are pulling about 10 times that much case vacuum. I was in a dyno room a few weeks ago with an engine builder who has measured a HP increase.

What was the magnitude of the power increase? Back of the envelope calculations say it would take a 6% increase in hp to deliver the 2% speed increase attributed to the revised crankcase vent.
 
What was the magnitude of the power increase? Back of the envelope calculations say it would take a 6% increase in hp to deliver the 2% speed increase attributed to the revised crankcase vent.

Which is why I suggested re-flying the speed test.
 
Crankcase Vent, in/out cowl, speed tests

Finally had a chance to perform the tests.

This first test is with the vent in the cowl, roughly positioned in the same general area, but due to hose/pipe/interference, I could not get it in the same place. Of course the 2nd test was with the CC Vent tube back outside according to the previous pics.

Note: Calculations using spreadsheet from link posted by Dan H. earlier this thread.

CC VENT INSIDE:
GPS1:,TRK1:,GPS2:,TRK2:,GPS3:,TRK3:
178,347,181,254,178,163
Calculated vTRU: 178.1

CC VENT OUTSIDE:
GPS1:,TRK1:,GPS2:,TRK2:,GPS3:,TRK3:
178,347,181,254,181,163
Calculated vTRU:179.6

Conditions:
OAT: 72F (at altitude, from Dynon)
BARO:30.29 (@6250 MSL, KTVL)
MSL: 7520 (with 30.29)
PA: 7150 (for both tests)
RPM: 2610
MP: 23.4
AVG IAS: 148-149 (yes, know it's way off, not using Van's rivets on static ports)
AVG ITRU: 172-173 (from Dynon)
Winds: light
Estimated Fuel Start: 35-36 gallons (+ or - 1 gallon)
Estimated Fuel End: 30-31 gallons (+ or - 1 gallon)
Flight duration (each): .3 Hobbs (.1 taxi)

Both tests were flown within ~1 hour of each other, over the same course, but there's no guarantee the wind conditions were exactly the same during the two runs.

Michael
 
When we do flight testing were the speed delta is going to be this small, we never consider the result valid without using data that was proven to be repeatable ( each test condition repeated a number of times with matching result. )

One way to help balance the test condition between each test interval is to use the density altitude feature that is on most EFIS systems, and do each test at the same density altitude.
 
Good to know, thanks. If I have a chance, I will do some further tests to see if the results are repeatable.
 
Scott is right of course.

Note that experimenters running exhaust evacuators are pulling roughly 10x the case vacuum report here.

A clean belly is a good result. Keep on eye on things, report back please.
 
Question regarding Density Altitude

I didn't record DA during those flights, but I *should* be able to calculate using a calculator such as: https://wahiduddin.net/calc/calc_da.htm

I also didn't have the dewpoint recorded for KTVL; however, the data is available from the NWS: http://w1.weather.gov/obhistory/KTVL.html.

Given the above, and that during my two flights were between 09:00 - 10:00 on 6/27/2015, the reported dewpoints (@6250 MSL - airport elevation) were:

1st flight: 42F (calculated DA: 9715)
2nd flight: 47F (calculated DA: 9745)

Given the above, should the results from the 1st & 2nd test just be thrown out?

(Since the calculated DA was not the same between the flights? -- 30 foot difference)

Or, can I salvage this test in any reasonable way?

Maybe the question is: Is it even possible for somebody that's not a professional test pilot to gather meaningful/valid flight performance data?
 
Down the gear leg?

What not run a breather line down the gear leg fairing(inside) and out the back of the trailing edge 2/3 of the way down the gear leg? A pitot shaped exit , but out the trailing edge pointing backward for suction. You would still need a whistle slot inside the cowl for safety, if the breather exit blocked or froze over.

I find if I run 8 Quarts in a 12 quart 540 Sump, there is minimal oil goes out the breather, so It is really a non issue for me, for most of my flying.

If I fly aerobatics, then I get oil streaks no matter what. Some of the hard core aerobatic AC use a breather line all the way to the tail wheel for an exit.

I think the gear leg would be less plumbing, if it could be made to work.
 
Maybe the question is: Is it even possible for somebody that's not a professional test pilot to gather meaningful/valid flight performance data?

Most definitely, yes, nothing magical about professional test pilots. But, one must come to terms with the fact that any one estimate of airspeed is measured with "error", where error is defined as random variation among replicate measurements. Some of this error is real variation among flights due to differences in bugs on the wings, turbulence, c.g., etc. Some is due to measurement error, for example, nonstandard lapse rate. If you follow the NTPS protocol (GPS track and ground speed) and use a modern EFIS to record data, you can minimize but not eliminate measurement error. In order to have any real confidence in your results, you need to conduct repeated tests, both before and after modification, and then evaluate the mean change in performance, pre. vs. post, relative to variation among replicate flights - a statistical test for change in means. If you find a significant change, great. If you don't, well, you may still have improved performance, but the magnitude of performance improvement falls within the range of the measurement error.

There are tons of posts here over the years where someone made a modification, and them made a single test run to evaluate its effect. Such tests are simply meaningless without a knowledge of error. Replication is how we estimate error.
 
Last edited:
Back
Top