[erich weaver]

Did y'all take note of the article on baffles by walter atkinson in the March Kitplanes issue? He and George Braly of GAMI investigated airflow conditions under the cowl by placing numerous tell-tales on the engine and installing a video camera. They got surprising results, some quite contrary to conventional wisdom:

1 airflow goes from the inlet around the primeter of the engine to the aft baffle, then to the center and back up to the front, where it exits the intake!

2 when running stationary on the ground, cooling is better when the plane is pointed DOWNWIND!

3 "Using RTV...is a waste of time..."

Check it out. Good stuff Walter! The article goes further into details, although Im still trying to figure out how much of the rest I can directly apply to my particular engine. I got first dibs on having Walter look over my installation and provide some tips.

erich weaver

 

[osxuser]

I believe it about rtv, but it still looks better .

 

[rvbuilder2002]

I read the article.
I believe what he wrote about the particular airplane he was testing is true.
What I am highly sceptical of is that we can automatically apply this to all airplanes.
There are many different flavors of baffle/plenum systems. Different inlet exit styles, etc.
For example, an RV with properly installed baffling has no air path from the plenum to the back of the spinner...so there can be no air leakage/flow out around the spinner (at least not air from the upper plenum).

 

[cjensen]

I found the article very well written and informative, but like Scott said, this can really only be applied to the particular airplane they were testing. Without doing this to every combination out there, we simply can't draw conclusions that this is fact for each type.

Very surprising though! I copied it, and showed it to all of our A&P's, and they were all enlightened by it as well, but had the same response. Great article though!

 

[erich weaver]

The article reminded me of a strange thing I noticed with the leaf blower I use at home. Like every leaf blower, it blows air out the end of a tube at high speed - 200 mph if you believe the manufacturers claims. The weird thing is that if you manage to get a leaf right located in just the right spot in directly in front of the tube, it gets SUCKED INTO the tube instead of getting blown away. Totally opposite air flow than expected, just like in Walter's article, where air flowed OUT of the intake as well as in.

One of the points I inferred from the article was that the volume of air coming in is so high that you really dont need to worry about sealing up every little leak point. Walter also says that the intakes should be "as small as possible", and I assumed that this is to avoid the tendency for the air to circuate inside the cowling and exit backout the front. Our standard RV cowlings seem to have reatively large inlets - has me wondering if they arent too large.

Walter, if you are out there, any additional specifics you can offer with respect to the RV baffles?

erich weaver

 

[rv969wf]

Reversed air flow is a NO NO

Some of the testing on cooling issues that I've found is somewhat true with the article, but every plenum and cowling design is different from one airplane to the other. For example,,, from the testing that I did on my -6 with many mods to the Plenum and Air Inlets to cool the engine,,,,I found out that the engine ran hotter with 4.125" round inlets per side than with 2.750" inlets per side. I have many post on VAF with some of this and it all has to do with air management and keeping the air movement where it needs to go and perfecting it.

It's all about air management and letting the air stay attached to the places it needs to go and making use of it 100%

 

[DanH]

Chris Zavatson (Lancair 91CZ) has some interesting technical papers posted on his website. #3405 is a classic for flat cowls.

http://www.n91cz.com/Interesting_Technical_Reports/Technical_Papers.htm

 

[frankh]

Some of the testing on cooling issues that I've found is somewhat true with the article, but every plenum and cowling design is different from one airplane to the other. For example,,, from the testing that I did on my -6 with many mods to the Plenum and Air Inlets to cool the engine,,,,I found out that the engine ran hotter with 4.125" round inlets per side than with 2.750" inlets per side. I have many post on VAF with some of this and it all has to do with air management and keeping the air movement where it needs to go and perfecting it.

It's all about air management and letting the air stay attached to the places it needs to go and making use of it 100%

One day I'm going to start experimenting with my Sam James cowl and the smaller inlets is item #1.

I have a relatively hot #4 cylinder that I havent gotten to the bottom of yet, still have to balnce my injectors, inish my instrument rating, take the Wife on the long trip she has been waiting for 2 years for...Get the lawn tractor running...and.......

Where does the time go?

 

[hevansrv7a]

Tweaking cooling in SJ Cowl/Plenum

One day I'm going to start experimenting with my Sam James cowl and the smaller inlets is item #1.

I have a relatively hot #4 cylinder that I havent gotten to the bottom of yet, still have to balnce my injectors, inish my instrument rating, take the Wife on the long trip she has been waiting for 2 years for...Get the lawn tractor running...and.......

Where does the time go?

With my Superior IO-360 Plus and the SJ setup, the two on the left were almost exactly the same and the two on the right were (front) lower and (back) higher. Total spread was 25-30 degF. The average of the right side was about equal to the two on the left. Sam said to make the lower inter-cylinder gaps as small as he recommended in his plans (I haven't, yet, because I used Van's baffle kit, cut down). I added a little vertically to the air dam in front of the right front cylinder (#1). That evened out the right side temps to a spread of less than 20 degF. I am using a small hole behind #4 for the heater air supply, too. These are all CHT degrees. Sam explained that with #1 being forward of #2, the cooling is a little more effective, so that's what I was trying to modify. So far, it's working quite well. I've never seen CHT above 375 in any configuration on this 'plane. I hope this is helpful. Sam is probably right about the lower gaps, too, but that's a lot more work, now.

 

[N91CZ]

Scott and Chad hit the nail on the head. The conclusions presented in the article are indeed tied to the aircraft used. Much of their testing was done on a Beechcraft Bonanza with a 6 cylinder Continental.
The RTV comment is particularly amusing and unfortunately misleading when applied to all other aircraft but to homebuilts in particular. GA cooling systems were designed to work with big leaks included. They flow far more air than is needed. That is different than saying they have excess cooling. They just flow more air than is needed. Some of the faster homebuilt designs tried to help performance by reducing cooling drag -that is by limiting the amount of air pumped through the engine compartment. Now leaks become much more critical. NASA put out a great report where this was actually measured. (http://www.n91cz.com/Interesting_Technical_Reports/NASA_CR3405.pdf) 17% was leaking past baffle to engine interfaces. RTV recaptured this loss completely. Will you notice the difference? Let?s say you designed the optimal cooling system that flowed exactly the right amount of air needed to obtain your target CHTs. In this case, any leak at all will be noticed. Loosing 17% of it will be a disaster. On the other hand, let?s say your cooling system was designed to flow 50% more air than you need. Well then fixing leaks here and there with RTV really won?t stand out that much. In the end, it depends on your point of reference.

 

[chuck]

I found the article very well written and informative, but like Scott said, this can really only be applied to the particular airplane they were testing. Without doing this to every combination out there, we simply can't draw conclusions that this is fact for each type.

I couple of posts made the point that Walter's findings aren't applicable to all types. That is true but I think that is not the point of understanding this. What is applicable to all combinations is that air does not do what most of us think. If you don't measure it (or visualize it), you will likely get it wrong because most of us have simple mental models that are not adequate for the complex flow.

 

[mistergerf]

I recently had a minor oil leak in the lower pressure side of the engine - the splatter came OUT of the cooling intakes at the front of the cowling and splattered along the top of the cowling and over the windscreen. I think this was one of the observations listed in the article.

g

(RV-9, IO-320 D1A, 50 hours)

 

[atreff]

Old Mooney cooling drag discovery

I have had similar experiences in another 'store bought' brand.

My 1966 Mooney C model was built with a very large cooling lnlet all the way around the prop. Huge. Additionally the oil cooler is mounted in the lower fwd cowl right out front, with ram air passing directly into the vanes.

One day, I noticed an oil streak along the outside of the lower cowl. It was coming from the lower edge of the cowl cutout for the oil cooler, down on the chin of the cowl. This position is really as far fwd as you can get, so I assumed I had a leaky cooler. I opened it up, and cooler and hoses were dry, but the inside bottom of the cowl was moist.

I couldn't see a leak anywhere. Washed the engine, flew it and finally found the leak. It was coming from a small fitting in the accessory case, running down the rear of the access case, and when it hit the firewall, the airflow pushed it forward all the way up front (under the airflow coming down from the cylinders) and uphill on the inner surface of the 'chin' of the cowl then out the cooler opening. Total distance travelled fwd? Approx 30" that slopes upward around 3", and the final climb is another near vertical 3" climb!

I was shocked that the air was making a 180 degree turn and going uphill! That made me realize why the cowl flaps were ineffective in the air. A significant amount of the air was exiting the oil cooler opening up front!

This configuration yielded oil temps consistently high-always 200-240. And CHT was very low.

I believe the cooling opening was waaaaay too large for the job at hand (O-360) in a 140 Kts TAS airplane. How do I know? I bought an STC'd kit that reduced the inlet by approx 70%. I picked up approx 7 kts TAS. Oil temps were now the opposite: very low 160-180 (I now have the oil cooler half covered to elevate oil temp), and the CHT's did the opposite as well, in climb, 400 degrees, then in cruise 350-375.

And for all of us Sam James Plenum Chamber fans out there, this Mooney was manufactured with an all alum plenum chamber back in 1966, and it still had rampant cooling drag due to Mooney's overly large inlet configuration. So, just having a plenum is not a panacea, everything must be in balance to achieve cooling drag nirvana.

AS to the Kitplanes article, every FWF configuration has it's own particular quirks. I would avoid any broad sweeping statements about all baffling for all planes.

And RTV? Totally necessary now, with my cowl inlets closed down. Case in point: I changed the engine, and put the Mooney stock plenum back on, and ran her up. CHT's were thru the roof. RTV'd the interface between the top of the accessory case and the plenum's 'back wall', and the temps came back down. Dramatically.

The only sweeping statement I'll make on baffling/cooling drag is this:

Chris Zavatson 's website (www.n91cz) has the most scientific instrumentation of cooling drag I have seen anywhere, he measures pressure in the cowl in various places as well as airspeed of incoming cooling air, and he teaches you how to do it in your plane! And.......he's good guy.....(for a Lancair guy),. If you are a student of cooling drag reduction, I'd go there and read his excellent paper on the subject. Could save you hundreds of hours of trial and error.

Chris Z-we'v not met, but I'm an old friend of Trey's who turned me onto your site and your award winning ship. Impressive

Art in Asheville

 

[rvbruce]

Baffle air flow

I find it interesting that the air starts working from the back forward. Not surprised though. If you have ever noticed a (round) parachute opening after deployment, you will notice that it opens from the TOP downward. Air is forced up thru the center channel and billows out the top first and as pressure builds it forces out the skirt to an open canopy.