What's new
Van's Air Force

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

High CHTs

On the mounting screw. Here the screw point is just above my left index finger.

Of course, the washer spaces the tin away from the head at the black rectangle as well as the no-fin depth area in the red circle. Airflow at the black rectangle is just a leak. That's why I referred you to how to make a bypass duct instead.


OK that explains the right rear (#5 in my case). What's the solution for the left rear (#6) as there's no onto cylinder mounting screw used on that side. I'm just looking for a temp setup to run some tests with the idea that the permanent solution are the ducts.
 
OK that explains the right rear (#5 in my case). What's the solution for the left rear (#6) as there's no onto cylinder mounting screw used on that side. I'm just looking for a temp setup to run some tests with the idea that the permanent solution are the ducts.

Look at your head - that restriction is only on one side. So - -it is only front left and aft right head that is affected by the baffles.
 
Yes Dan, but... if the cut only slants forward an inch, what is the speed loss? 2 knots, maybe 3? Well worth the tradeoff, if all else fails. Also, IL looks better than adding louvers or a lip and simpler than adding a cowl flap.

Bill R. asked an interesting question. Although it's hard to precisely quantify in terms of speed loss, it can be quantified in terms of exit area increase. Knowing how much you actually opened the exit is important, because it offers some realistic expectation about additional cooling.

For five angles of cut (see the diagram below) the multiples of original exit area are:

10 deg 1.0154
20 deg 1.0642
30 deg 1.1547
40 deg 1.3054
45 deg 1.4142

What this little trig exercise tells us is that you'll need some significant cut angle to actually increase area very much. For example, assume the existing RV-6 cowl exit is 6" high (I'm guessing, don't have one handy). Cutting it back 1" would be a cut angle a fuzz less than 10 degrees, so the area multiplier would be 1.0154, or only about 1.5% larger than stock...which would not make any measurable difference in speed or cooling.

IF you elect to cut the exit for more area, you'll probably want to choose 30, 40, or 45 degree cuts. For an exit originally 6" high, those angles would mean cutting forward roughly 3.5", 5", or 6", for an area increase of 15%, 30%, or 41% respectively.
 
Last edited:
Here's what I have done so far:

I removed the 2006 oil cooler and fitted the larger 2007. I then made an L-shaped plate which I temporarily fitted with RTV over the oil cooler outlet in the rear baffle. I covered about 30% of the hole and bent the "L" to about 45 deg so as to guide (hopefully) air over the No6 cylinder.

This was partially successful. The oil temperature remains normal. The No6 CHT dropped by about 15F but that still leaves it 15-20F above the No5 which is the next highest (that is about 15-20 higher than the front 4 cylinders). The boroscope on my annual check confirmed that No6 was running a little warmer but everything looked normal.

So, if I go full chat in race trim and lean to around 19gph (about best power), the number 5 gets to just over 400 and the No6 goes to about 420. I'm not going to do that! I need 21gph to keep the number 6 at around 400 - which I can live with.

I am going to take the cowls off again and check once more the sealing on the lower baffles but I am convinced there is nothing wrong on the top side.

I have one more thought. Would it help to put a slightly larger injector on the No6? I cruise LOP and have a GAMI spread around 0.3 and other than going flat out, my CHTs are fine even if No6 is a little higher than I would like - so I am not sure I actually want to screw with any of this.
 
My current rv10 will climb full throttle, 90kts all day and never hit 400. Bone stock cowl, baffles, louvers. Gear slot extended maybe 2 1/2 inches. They should all do this. If one can't, then something is being missed.

Actual Repeat Offendrer, I would very much like to see some photos of your baffles installtion next time you open the cowl. Although, I don?t have the same issue, I do go over 400 deg in climb, summer of winter. Thanks.
 
I struggle at times in the summer heat with CHTs that will hit 400 in the climb if I don?t reduce power and keep the airspeed up to 125. Most recently I made a mod to the heating system which helped significantly. I blocked off the right side 2? exit at the rear of the baffle that supplies air to the right heat muff. I then put a splitter in the line coming from the left side forward air supply for the left heat muff and routed it to the right side. My CHTs have come down around 10 degrees with this mod.
There is still more than enough heat coming in with this mod
 
I might try that - I already have the front intake feeding both hot air outlets. I have kept the rear baffle vent connected for potential use as a vent to keep tunnel temps down if required. However, there doesn't seem to be an issue there.
 
One thing I can't agree with is that the oil duct doesn't "steal" air. The argument that the pressure equalizes I don't think holds water. Since the air is a moving compressible fluid, the pressure is going to be different all over the baffles, so air going through the duct can for sure cause a lower pressure (and therefore less cooling) over the no 6 cylinder.

Returning to a thread from a few months ago....

It's a common belief that an oil cooler inlet somehow steals air from the adjacent rear cylinder, causing it to exhibit a high CHT. I've suggested it does not do so, based on some previous observation of CHT with changes in oil cooler air supply location. I've also said that measuring static pressure at various points within the upper plenum space would tell a lot.

As it happens, I recently rigged my plenum with pressure taps at each FI nozzle location, a story for another time. In the context of stolen air, I thought everyone might like to see the reality.

Below you can see the pressure taps on the #1 and #3 nozzles. They are simple aquarium bubble rocks, the goal being to record static pressure, not dynamic. The taps on #2 and #4 are similar, but note they are on the forward facing part of the head, where the conversion of dynamic pressure to static is a little higher, rather than on the rear facing part like 1 and 3.

Key point here is the location of the #3 tap, immediately in front of the oil cooler duct inlet:

Oil%20Cooler%20Iinlet%20Pressure%20Loss.jpg


And the numbers, at three altitudes, with three speeds at each altitude:

Pressure%20Tap%20Data.jpg


Note that #3 pressures are not significantly different, and in fact are just a bit higher than those recorded for the #1 nozzle well forward of the oil cooler intake.
 
Last edited:
Dan, I agree with your conclusion, but I’m not sure your pressure taps prove the point. It seems you should have bottom deck taps as well, so you can see the pressure drop over each cylinder and over the oil cooler and, perhaps, test both with and without the oil cooler path blocked.

The notion that the oil cooler path “steals” air from the adjacent cylinder is presumably based on a flawed presumption that there is a static supply of cooling air (i.e., if it didn’t go over the cooler it would have gone over #3). In reality, the cooling air will travel through the path of least resistance, which appears to many as that 4” scat tube behind #3, but if one takes that simplistic approach to its logical conclusion, he’d convince himself that NONE of the available cooling air goes over the cylinders. Obviously, the scat tube isn’t the end of the story. The air must get through the oil cooler and then out the exit area (along with the other air in the exit area that came over the cylinders or through baffle leaks).

Your pressure readings show us that the static pressure in the top deck is reasonably uniform regardless of the location. What they don’t show is the pressure drop over each cylinder and the oil cooler. The pressure drop at each location is directly related to the speed and volume of cooling air, so it seems that would be better evidence to support your conclusion, particularly if shown with the oil cooler path both open and blocked.
 
Your pressure readings show us that the static pressure in the top deck is reasonably uniform regardless of the location.

Correct.

What they don?t show is the pressure drop over each cylinder and the oil cooler.

Cylinder and cooler delta would be useless in this context. A second round of upper deck measurements with the cooler duct entrance blocked might be illustrative.
 
A second test with the cooler path blocked would be interesting but I think the numbers you have already show that the pressures are about equal around the cylinder with the cooler duct near by. Blocking the cooler path may/will increase the pressure in the upper deck but it appears that it would be an equal increase throughout the upper deck.
Meaning that the cooler duct behind/ above #3 does not decrease airflow to #3. Therefore would not increase the CHT to #3 cyl.

Not being argumentative. Just pondering.
 
Additional points to ponder, if you will...

Does the oil cooler itself represent a greater or lesser restriction to airflow as compared to the flow restriction represented by the #3 cylinder? We tend to work with the assumption that tapping off a 3" SCAT above the #3 cylinder is equivalent to a total loss of pressure in that area, whereas any resistance to flow offered by the oil cooler would perhaps reduce the pressure drop above Cylinder #3 to something equivalent to perhaps a 1.5" or 2" free-flowing SCAT tube. This is far more a question than a statement - please educate me on this point.
 
Does the oil cooler itself represent a greater or lesser restriction to airflow as compared to the flow restriction represented by the #3 cylinder?

That's an interesting question. Thinking out loud...lets use the pressure delta I posted here...

http://www.vansairforce.com/community/showthread.php?p=1169174

...which was about 7" H20 at 150 KTAS.

If we go to a Lycoming cooling chart for my 390...

390%20Cooling%20Chart.jpg


....we can see that 7" H2O flows 2.4 lbs/sec at sea level, or 144 lbs/min. Remember, that's for all cylinders, and it is empirical...it assumes all users have some "standard baffle" setup like whatever they use in the dyno room in Williamsport.

Now go to the Stewart Warner data, and pull up the chart for the SW10611 I'm using....

SW%2010611%20Air%20Side%20Drop.jpg


...and we see a delta of 7" H2O flows 47 lbs/min (note, at 0.0765 lbs per cubic foot, which is why I used sea level on the previous chart). Here we must remember the flow assumes a bare cooler, with no duct losses. Given the same 7" drop, a scat tube would mean fewer lbs per minute.

So how does a cooler compare to one cylinder? In ballpark terms, we have 144/4, or 36 lbs/min for a single cylinder's baffling. The bare cooler is 47 lbs/min, but the duct would reduce that value. All in all, it looks like this particular cooler, when fed with a duct, offers resistance to flow very similar to a "standard" cylinder, and thus would not be a pressure sink.

My baffling is probably a lot tighter than the Lycoming example, which would push the 36 lbs/min per cylinder to some lower value. I also duct my cooler exit all the way to a point in the cowl exit bell, in an attempt to tap pressure lower than that found in the lower plenum, which should push cooler mass flow upward. So, mine should tend toward being a sink, if such a thing is true, but it does not seem to be the case.
 
Last edited:
A second test with the cooler path blocked would be interesting but I think the numbers you have already show that the pressures are about equal around the cylinder with the cooler duct near by. Blocking the cooler path may/will increase the pressure in the upper deck but it appears that it would be an equal increase throughout the upper deck.
Meaning that the cooler duct behind/ above #3 does not decrease airflow to #3. Therefore would not increase the CHT to #3 cyl.

Not being argumentative. Just pondering.

Mark, the cooling is a "system" consisting of leaks, pressure drops for each component, inlet sizes, etc. just to mention a few. But - on my system during gathering of upper plenum pressure measurements, there is no change in the pressure in the upper chamber between full open and fully closed oil cooler shutter. So, while someone's system maybe different, this is one example where the assumption was not validated with data. I have practically zero pressure (compared to the static port) in the lower cowl chamber. Steady state cruise during the testing had no change in CHT's, no trend, none.

Let me add, regarding a change in the airflow across the heads - the pressure drop is the same so unlikely, also - the inlets are certainly not choked allowing a drop in plenum pressure. I can only conclude that the total mass flow is increased to handle the flow across the cooler making me happy with my design. BTW this system uses the standard Niagara Vans cooler on the baffle.

Edit: (echoing DanH comments below) This is an example of the system at work, nothing here should be taken at rote. This is an example (an aid) to better understand "standard fixes". It is on the other end of the system scale, with very tight baffling and controlled air across the engine fins. Certainly not a well balanced system yet. I've thought about adding a large throttled bypass just to change the mass flow in flight for assessing/quantifying engine/cooler delta-p effects. In fact, I'm not even sure what the "well balanced" criteria should be. I'd love a round table discussion about that some day.
 
Last edited:
....on my system during gathering of upper plenum pressure measurements, there is no change in the pressure in the upper chamber between full open and fully closed oil cooler shutter.

Bill's installation is baffled very tightly, and incorporates carefully shaped internal intake expansion. For sure it is not typical.

Based on preliminary data, it appears the engine baffles are currently the system throttle, rather than the cowl exit, which is too large for the flow. To wit.....

I have practically zero pressure (compared to the static port) in the lower cowl chamber.

It's an application of the old inlet-exit ratio some designers insisted on (exit larger than the inlet), but Bill has turned it on its head with tight baffling and excellent pressure recovery.

I can only conclude that the total mass flow is increased to handle the flow across the cooler making me happy with my design. BTW this system uses the standard Niagara Vans cooler on the baffle.

Couple interesting things here, although I fear much confusion may be found in exploring them.

Regardless of the location of the system's throttle, mass flow does increase when the throttle is opened.

Here the variable exit device is the oil cooler shutter. The little Niagara is a small cooler, and the shutter makes the available flow area even smaller. I'm pretty sure it would not cool adequately given a typical pressure delta, but Bill's deltas are abnormally high (for example, 11+ inches H20 at 150 knots), so it works. It may not allow shrinking the exit, as the delta will decrease.

In a typical system, throttling the exit would result in an upper deck pressure rise. I suspect there is a pressure rise with Bill's system too, like any other, but it is small enough to be lost in the noise. The change in exit area is small, and the manometer display bounces around because of prop outflow. It's not a squirrel cage fan ;)

There may be one other factor in play, but I'd need to do some reading first. For now, just be careful what lessons you take from Bill's system, because it's not like yours.
 
Last edited:
How does static pressure at the nozzles effect fuel atomization?

Sorry for the thread drift but I am trying to reason through an air pressure difference at the nozzle vent screens and what that might do to fuel atomization and how it burns. Perhaps a new thread is in order. In your case they are quite close, but I suspect a larger spread can be found in other cowling examples.

Returning to a thread from a few months ago....

It's a common belief that an oil cooler inlet somehow steals air from the adjacent rear cylinder, causing it to exhibit a high CHT. I've suggested it does not do so, based on some previous observation of CHT with changes in oil cooler air supply location. I've also said that measuring static pressure at various points within the upper plenum space would tell a lot.

As it happens, I recently rigged my plenum with pressure taps at each FI nozzle location, a story for another time. In the context of stolen air, I thought everyone might like to see the reality.

Below you can see the pressure taps on the #1 and #3 nozzles. They are simple aquarium bubble rocks, the goal being to record static pressure, not dynamic. The taps on #2 and #4 are similar, but note they are on the forward facing part of the head, where the conversion of dynamic pressure to static is a little higher, rather than on the rear facing part like 1 and 3.

Key point here is the location of the #3 tap, immediately in front of the oil cooler duct inlet:

Oil%20Cooler%20Iinlet%20Pressure%20Loss.jpg


And the numbers, at three altitudes, with three speeds at each altitude:

Pressure%20Tap%20Data.jpg


Note that #3 pressures are not significantly different, and in fact are just a bit higher than those recorded for the #1 nozzle well forward of the oil cooler intake.
 
Somewhat relevant I hope.....

I had the oil cooler in the stock position behind cylinder #4, I thought it was stealing air from #4 cylinder and preheating the air to the oil cooler. I made a remote oil cooler via a rubber inlet duct and a fiberglass exit. In detail:

http://www.vansairforce.com/community/showthread.php?t=94648&page=53

IMG_5751-L.jpg


I did see temp improvements on Cylinder #4, I believe the lower temps to be attributed to the curvy baffling additions and not a pressure drop from the oil cooler stealing air.

IMG_5673-L.jpg


Oil temps went up, seems my oil cooler inlet/boot isn't very efficient losing several inches of H2O from the upper plenum pressure.

What surprised me the most is my oil temps increase in a descent. Pulling power and nosing over as my airspeed increases CHT's start dropping and oil temp rises. Since my lower cowling is throttled down to 35ish sq inches both my upper and lower cowling pressures increase. I believe this happens because my oil cooler duct only converts a percentage of the upper plenum pressure to pressure at the oil cooler face, the pressure differential across it decreases with the higher lower cowling pressures in the descent.

I'm hoping this winter to reattach my oil cooler to my baffling with a short duct to re-capture full plenum pressure. I plan on leaving the curvy #4 baffling additions on.
 
Last edited:
Late to the party

Here's what I have done so far:

I removed the 2006 oil cooler and fitted the larger 2007. I then made an L-shaped plate which I temporarily fitted with RTV over the oil cooler outlet in the rear baffle. I covered about 30% of the hole and bent the "L" to about 45 deg so as to guide (hopefully) air over the No6 cylinder.

This was partially successful. The oil temperature remains normal. The No6 CHT dropped by about 15F but that still leaves it 15-20F above the No5 which is the next highest (that is about 15-20 higher than the front 4 cylinders). The boroscope on my annual check confirmed that No6 was running a little warmer but everything looked normal.

So, if I go full chat in race trim and lean to around 19gph (about best power), the number 5 gets to just over 400 and the No6 goes to about 420. I'm not going to do that! I need 21gph to keep the number 6 at around 400 - which I can live with.

I am going to take the cowls off again and check once more the sealing on the lower baffles but I am convinced there is nothing wrong on the top side.

I have one more thought. Would it help to put a slightly larger injector on the No6? I cruise LOP and have a GAMI spread around 0.3 and other than going flat out, my CHTs are fine even if No6 is a little higher than I would like - so I am not sure I actually want to screw with any of this.

Why not put a proper bypass plenum on Cylinder #5 and drop your temps to where it is the coolest cylinder
( see Dans back RH cylinder below) This has been proven by multiple builders to provide 2x the temp reduction to that cylinder vs the washer method that wastes and flows air over useless non finned areas.
Screen-Shot-2018-09-08-at-8.15.11-AM-e1536416439902.png
,
Then move your oil cooler intake to that side, #5 where you have more cylinder cooling than needed. On the other side, you will have all of your available air flow(perceived or real) to Cylinder #6, with no loss to your oil cooler.
 
Last edited:
Sorry for the thread drift but I am trying to reason through an air pressure difference at the nozzle vent screens and what that might do to fuel atomization and how it burns.

Looking into it right now. Short version; there is a huge bleed air delta at low MP, and very little at high MP. Luckily, we need at lot of help with atomization at low MP, and not so much at high MP. More later, different thread.
 
Back
Top