What's new
Van's Air Force

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

High CHT's resolved and what worked.

Paul 5r4

Well Known Member
Hi Friends...
History. I've been flying the RV 7A for two years. On climb out the number two CHT would always hit 400 before I reached 1000 feet! I had learned to live with it doing a kind of step climb thing.... level off/thottle back/ cool/then climb some more or enter a shallow cruise climb. I had tried the usual suggestions tight baffles etc. without success. Here's what worked and worked GREAT with several pics.

You MUST/HAVE TO/WITHOUT FAIL.... get airflow beneath the forward part of the number two cylinder. I cut a hole then built a channel for air to get beneath that cylinder. Success! Now I can climb to around 4000 feet before I hit the upper 390? range. Around 4000 feet, the temperature levels off then starts dropping because of the cooler OAT's. Having such great success with the number two cylinder I decided to attack the number three CHT the same way. It didn't get quite as hot as number two but it was pretty close. Yesterday just for fun I took off and climbed at 120 kts all the way to 10,500. Incidentally, I was heading into the wind during by climb to 10,500 and my groundspeed had dropped to 99 mph! When I turned around to head home I was doing 240 mph across the ground! Wish I could do that every time I fly!

Straight up... If a rather sloppy RTV application offends you, better bail here.
Otherwise, here are the pis of the CHT fix. (I had reached the point that I was getting ready to start cutting the cowl to increase exit airflow. Happy now I had not done that yet!!!)

You can visualize the airflow to the fins
2ztc2ld.jpg

Number two cylinder from the bottom of the baffle
ndsvhj.jpg

Behind number three cylinder. Note minimal clearance for airflow
s4xk6x.jpg

fcqvdh.jpg

Hole cut. I almost cut the wrong section of the baffle
30dibr6.jpg

From the back
9qig6t.jpg

Front/inside of baffle
2lw62yw.jpg

Looking down through newly created airflow channel
312hfz5.jpg

From the back behind number three cylinder
2ai264z.jpg
 
Last edited:
Right answer, wrong physics

Around 4000 feet, the temperature levels off then starts dropping because of the cooler OAT's.

The CHTs start dropping because the engine is putting out less power. True, as you go up, OATs drop - but so does air density. Air density dominates. If the engine could put out constant power you would see CHTs rise with altitude.
 
Dan, yours is a perfect and elegant solution. I wish I would have done mine that way. Sure looks better. End result is the same though and I'm thankful it worked.

Bob, Yes the engine produces less power with increasing altitude.... and truthfully I didn't think of that. With that said, as I previously mentioned I couldn't get to 1000 feet without hitting 400? and the temperature was still heading well north of that 400 mark as was cylinder no. three. I successfully brought my cylinder head tips down to normal levels in the first 4000 feet of climb... I'll let physics take care of the rest and thank you for helping me understand better. :)
 
The CHTs start dropping because the engine is putting out less power. True, as you go up, OATs drop - but so does air density. Air density dominates. If the engine could put out constant power you would see CHTs rise with altitude.

Air density is indeed dropping, which is the same as saying that available cooling mass flow is also dropping.
 
The CHTs start dropping because the engine is putting out less power. True, as you go up, OATs drop - but so does air density. Air density dominates. If the engine could put out constant power you would see CHTs rise with altitude.

Also true. CHT's can be the limiting factor in a turbo charged airplane going high. You get the same internal pressures and heat as sea level, with much less cooling mass.
 
The question I have for you is, are you sure you have changed the cylinder temperatures or just the temperature indication? You basically are pushing cooler air directly at the temp probes this can cause a change in indicated temperature, but it might not change actual cylinder temperature.
 
I will support the post above. My thoughts exactly.

Unless you take 6 probes placed all around the cylinder head and retest it you can be creating egg shaped Vs round.

This was done for the GAMI STC's for cooling on Bonanza cylinders, and nothing happens as you think it might, so hard data done with some proper analysis is the only way to know.
 
The question I have for you is, are you sure you have changed the cylinder temperatures or just the temperature indication? You basically are pushing cooler air directly at the temp probes this can cause a change in indicated temperature, but it might not change actual cylinder temperature.

Interesting supposition. I would reason that if the new airflow isn't changing cylinder head temperature, then it is certainly not harming anything.

Of course, if the airflow is having a effect on the cylinder head, how is it different from the existing airflow to the same area of cyls 1 and 4?
 
Last edited:
No biggie. The manufacturers have used that trick for years. "What do you mean the temps too high to certify? Oh well, put a blast tube on it. "
 
The Grumman Solution

No biggie. The manufacturers have used that trick for years. "What do you mean the temps too high to certify? Oh well, put a blast tube on it. "

Gumman solved the hi oil temp problem for my AA1A in the same manner - instead of an oil cooler (too heavy), the just put a bast tube onto the oil temp probe itself !
 
About twenty years ago, Airmotive Engineering was hired to resolve some cooling problems with a modified Mooney. It had high CHT?s in the climb and poor oil consumption. We heavily instrumented this aircraft and preformed extensive flight testing. One of the things we learned was there was a 100 F+ temperature difference between the top and bottom of all the cylinders on this aircraft. The exact same engine in our test cells showed a 0 F difference. With a large temperature difference the cylinders were not round any more. This can cause poor ring seal, blow-by, oil consumption and high CHT?s. When we corrected the cooling problems this engine showed a 5 F temperature difference, vastly improved oil consumption and normal temp control during the climb.

Also there are very hot areas of the head (around the exhaust) and relatively cool areas (around the intakes).This cool area around the intake does not need a large amount of air flow. That?s why there are smaller and fewer fins on the intake side of the head and a small regulated airflow. Heat always flows from the hot areas toward the cool areas, inside the head. The idea is to have a head that is as even in temperature as possible to help control internal thermal stresses.

Just because you haven?t observed or measured the changes in the heat flow path does not mean that no changes were made, good or bad. If this modification of a very well proven design with hundreds of thousands of successful operational hours, works so well don?t you think it would have been incorporated in all cooling systems before now.
I do not want to discourage anyone from trying new ideas. But it?s best when you can prove beyond a doubt, that you have accomplished what you think you have
 
When we corrected the cooling problems this engine showed a 5 F temperature difference...

Fascinating. What changes did you make?

Also there are very hot areas of the head (around the exhaust) and relatively cool areas (around the intakes).This cool area around the intake does not need a large amount of air flow. That’s why there are smaller and fewer fins on the intake side of the head and a small regulated airflow.

Again, cylinders 1 and 4 have unrestricted airflow to the same fins which are largely blocked on cylinders 2 and 3. Can you explain the difference?
 
Last edited:
"When we corrected the cooling problems this engine showed a 5 F temperature difference, vastly improved oil consumption and normal temp control ......"

Mr. Looper,

Could you please expand on this statement and share how you corrected the cooling problem?

Thank you
 
The engine was a Lycoming O-360A1A. Temperatures were measured on the barrel and head, top and bottom of all cylinders. This airframe had a custom made baffle system to go with a completely new cowling.
It had some poorly fitted lose fitting baffles in the area around the heads on all four corners and poor cooling pressure differential of 4?water. (It takes about 6-7?water to cool properly). One square inch of total leaking area in the baffling is equal to about 1?water cooling pressure.
When we corrected the cooling problems this engine showed a 5 F temperature difference. Between the top and bottom of the cylinders. Not between individual cylinders.
 
poor cooling pressure differential of 4?water. (It takes about 6-7?water to cool properly). One square inch of total leaking area in the baffling is equal to about 1?water cooling pressure.
Can anyone point me to a simple, preferably quick, means to check the pressure differential between top and bottom?
 
Can anyone point me to a simple, preferably quick, means to check the pressure differential between top and bottom?

You want quick and dirty? Pick up two aquarium bubble rocks, about 20 feet of 1/8 vinyl tubing, and 4 feet of 1/4" vinyl tubing. Ziptie one rock (with connecting tubing) to the top center seam of the engine case. Ziptie the other rock to the engine mount somewhere behind the engine. Neither should be positioned in a stream of high velocity air.

Run the pair of tubes back to the cockpit any way you can. Some have gone through the heater box. You could also run it out the cowl exit and up around the fuselage, and under a canopy edge. Duct tape it to the fuselage skin.

Fasten the 1/4" tubing to a 24" x 4" board so it forms a "U" with straight sides. Mark the board with lines straight across every half inch. Mix some food coloring and a bit of dish soap into a few ounces of water. Pour the water into the U-tube until it fills the bottom 10 inches. Now connect the engine compartment tubes to the ends of the U-tube.

Go fly. If the water jumps up and down too much, install restrictors in the lines....plugs with a 0.040" hole bored through them.

Pressure differential is the difference between the two water levels. Pressure varies with both altitude and airspeed. Email me for a Lycoming cooling air demand chart.
 
... If this modification of a very well proven design with hundreds of thousands of successful operational hours, works so well don?t you think it would have been incorporated in all cooling systems before now...

"This modification" addresses the significantly restricted/zero airflow at the point of zero fin depth on some Lycoming cylinders. Yes, there are hundreds of thousands of hours flying with restricted or zero airflow to the bottom fins, but are you suggesting this is an acceptable practice? Why go through the trouble of adding the wrap around baffle if there is no airflow?
 
"This modification" addresses the significantly restricted/zero airflow at the point of zero fin depth on some Lycoming cylinders. Yes, there are hundreds of thousands of hours flying with restricted or zero airflow to the bottom fins, but are you suggesting this is an acceptable practice? Why go through the trouble of adding the wrap around baffle if there is no airflow?

I am suggesting that this design has worked very well for over 50 years and it was purposely designed to have a small regulated airflow in this relatively cool part of the head (Front of # 2, rear of # 3 head area). Baffling in this area of the head should not be so tight fitting that it totally blocks the flow. That?s why most use a washer between the head and baffle where it attaches to the head. The only way it could have zero airflow, would be if this area was silicone shut.
Most Lycoming type engines and their clones go to TBO. They would not if they were not being cooled properly. This cooling design is well proven, something must be working right.
 
Most Lycoming type engines and their clones go to TBO. They would not if they were not being cooled properly. This cooling design is well proven, something must be working right.

The problem is, it is up to the airframe designer to create the baffles. Many airplanes have the baffling well below the mid point on the cylinder and allow plenty of airflow to the "bottom" fins. As we know with Vans baffles, the airflow IS essentially stopped up for the left fwd and right aft cylinders. The fact that washers are used as a crutch for this deficiency should be warning enough that the basic design is flawed. We also can assume that these bottom fins need airflow since they are quite large and are wrapped with a baffle.

It's obvious that the Lycoming is fairly tolerant of baffle design "sins", but we should not dismiss this tolerance as "successful". There are far more effective ways to cool a Lycoming than the kit supplied parts Vans puts out.
 
I am suggesting that this design has worked very well for over 50 years and it was purposely designed to have a small regulated airflow in this relatively cool part of the head (Front of # 2, rear of # 3 head area).

Quick sketch, just to illustrate the issue...

2923bba.jpg


Baffling in this area of the head should not be so tight fitting that it totally blocks the flow. That’s why most use a washer between the head and baffle where it attaches to the head. The only way it could have zero airflow, would be if this area was silicone shut.

Apparently you do feel some step must be taken to ensure "regulated airflow" to the fins below the restriction (red circle). A washer or two on the baffle mount bolt (the bolt hole is seen here below the black rectangle) does indeed space out the baffle, allowing lots of airflow through an area with no fins. That's a useless leak, fine for GA slugs with huge inlets and exit areas, but not so spiffy for efficient cooling.

sghoqu.jpg


Most Lycoming type engines and their clones go to TBO. They would not if they were not being cooled properly. This cooling design is well proven, something must be working right.

Have to agree. There's a pretty good argument for each cylinder being an individual engine. Temperature deltas between cylinders probably don't compromise TBO, assuming the hot one isn't too hot.

Break.

Bobby, could you please identify the source document for the pressure measurement drawing in the previous post? The baffle buttons are straightforward, but I can't quite tell what is intended with the Section A-A probes, or how they are mounted.

For those interested in such things, CR3405 made an excellent comparison of probe types and arrangements. See pages 14 through 20, and Figs 11, 12, and 13. The investigators preferred baffle buttons for upper plenum pressure and piccolo tubes for the lower plenum, as they found long piccolos in the upper plenum read slightly low compared to buttons. That's not a surprise, given that the buttons are not shielded against local flow velocity.

I chose foot-long piccolos for both upper and lower plenums, for several reasons. They are easy to make, mount, and plumb, in an identical fashion, important if global, easily comparable results are expected from a group of homebuilders. Two, they work pretty much the same regardless of plenum volume as they shrug off dynamic pressure. Three, they are self-averaging over the majority of the plenum volume.

http://www.vansairforce.com/community/showpost.php?p=880477&postcount=40
 
Last edited:
Cooling Issues

I have a SJ Cowl on Plennum on my -4, and have heating problems that require a step climb. In my research and efforts on trying to make things cooler I have noted the necessity to keep the differential (4" H2O). Wouldn't this fix as described in the opening to this discussion create more pressue in the lower area and thus restrict downward air flow?
My hotest cylinder is always #2.
Any other ideas for us SJ users?
 
Quick sketch, just to illustrate the issue...

2923bba.jpg




Apparently you do feel some step must be taken to ensure "regulated airflow" to the fins below the restriction (red circle). A washer or two on the baffle mount bolt (the bolt hole is seen here below the black rectangle) does indeed space out the baffle, allowing lots of airflow through an area with no fins. That's a useless leak, fine for GA slugs with huge inlets and exit areas, but not so spiffy for efficient cooling.

sghoqu.jpg




Have to agree. There's a pretty good argument for each cylinder being an individual engine. Temperature deltas between cylinders probably don't compromise TBO, assuming the hot one isn't too hot.

Break.

Bobby, could you please identify the source document for the pressure measurement drawing in the previous post? The baffle buttons are straightforward, but I can't quite tell what is intended with the Section A-A probes, or how they are mounted.

For those interested in such things, CR3405 made an excellent comparison of probe types and arrangements. See pages 14 through 20, and Figs 11, 12, and 13. The investigators preferred baffle buttons for upper plenum pressure and piccolo tubes for the lower plenum, as they found long piccolos in the upper plenum read slightly low compared to buttons. That's not a surprise, given that the buttons are not shielded against local flow velocity.

I chose foot-long piccolos for both upper and lower plenums, for several reasons. They are easy to make, mount, and plumb, in an identical fashion, important if global, easily comparable results are expected from a group of homebuilders. Two, they work pretty much the same regardless of plenum volume as they shrug off dynamic pressure. Three, they are self-averaging over the majority of the plenum volume.

http://www.vansairforce.com/community/showpost.php?p=880477&postcount=40
I am going to check this gap on my engine to see if there is any space but I though this type of fins are limited to a certain cylinders and not all.

My temps over all is not bad, in the 360-370 range ROP and summer time and the delta between the hottest and coolest is about 10F or less. But I am limited on climbs and would love to improve that.

My hottest cylinder is #3
 
Quick sketch, just to illustrate the issue...

2923bba.jpg




Apparently you do feel some step must be taken to ensure "regulated airflow" to the fins below the restriction (red circle). A washer or two on the baffle mount bolt (the bolt hole is seen here below the black rectangle) does indeed space out the baffle, allowing lots of airflow through an area with no fins. That's a useless leak, fine for GA slugs with huge inlets and exit areas, but not so spiffy for efficient cooling.

sghoqu.jpg




Have to agree. There's a pretty good argument for each cylinder being an individual engine. Temperature deltas between cylinders probably don't compromise TBO, assuming the hot one isn't too hot.

Break.

Bobby, could you please identify the source document for the pressure measurement drawing in the previous post? The baffle buttons are straightforward, but I can't quite tell what is intended with the Section A-A probes, or how they are mounted.

For those interested in such things, CR3405 made an excellent comparison of probe types and arrangements. See pages 14 through 20, and Figs 11, 12, and 13. The investigators preferred baffle buttons for upper plenum pressure and piccolo tubes for the lower plenum, as they found long piccolos in the upper plenum read slightly low compared to buttons. That's not a surprise, given that the buttons are not shielded against local flow velocity.

I chose foot-long piccolos for both upper and lower plenums, for several reasons. They are easy to make, mount, and plumb, in an identical fashion, important if global, easily comparable results are expected from a group of homebuilders. Two, they work pretty much the same regardless of plenum volume as they shrug off dynamic pressure. Three, they are self-averaging over the majority of the plenum volume.

http://www.vansairforce.com/community/showpost.php?p=880477&postcount=40

Section AA is a copper tube soldered to a flat piece of steel. It attaches to an intake manifold bolt to hold the tube up inside the fins in between the valves. I have had this drawing for so long I forget where it came from; it most likely is a Lycoming service instruction
 
...I am going to check this gap on my engine to see if there is any space but I though this type of fins are limited to a certain cylinders and not all...

It appears that some cylinder types have more fin depth in this area than others, but the simple fact remains that when cylinders are "paired", the "deep fins" on the exhaust side of one cylinder provide an ample flow path for both. However, when you have the "shallow" (intake side) fins up against a baffle (as with the left fwd and aft right), then this flow path is almost completely gone. Several of us are simply trying to replicate this flow path so the engine can uniformly cool as designed.
 
Last edited:
Section AA is a copper tube soldered to a flat piece of steel. It attaches to an intake manifold bolt to hold the tube up inside the fins in between the valves. I have had this drawing for so long I forget where it came from; it most likely is a Lycoming service instruction

Most likely from Installation Design for Engine Cooling, an old in-house Lycoming manual. I've requested a copy in the past with no success.

A vertical, open-end tube would measure total pressure, which more or less indicates the investigator was spending his days next to a dyno, i.e. vertical cooling flows. No surprise.

Fig 11-1 has a few problems as instrumentation for in-flight research. First, there is no provision for lower plenum pressure. The system as shown assumes current static in its measurement of pressure drop, which isn't true in any decent cowl. Let's assume there is some other drawing detailing a lower cowl pressure tap setup, as drop can be determined by comparing an upper cowl-static port measurement against a lower cowl-static port measurement.

The other drawback is a bit more subtle. When lower cowl exit area is increased, overall mass flow is increased. When exit area is decreased, mass flow is decreased. It's an intuitive result; reducing lower cowl pressure causes more air to flow through the engine fins.

Here's the measurement problem. Consider the case of increased exit area and more mass flow. The increased inlet ratio (inlet velocity must rise to accommodate the increased mass) almost always comes with loss of pressure recovery and internal drag loss, unless the inlet's internal shape is very good. Because a total pressure tube responds to dynamic pressure as well as static pressure, the higher internal velocities (which are quite likely turbulent flow) would tend to show higher pressure on the instrument, while a static pressure probe (like a picciolo tube) would tend to show lower pressure.

Take a good look at CR3405's Fig13. Probe type(s) 2, 4, and 5 (total pressure tubes) are all over the plot, as the individual locations are subject to different local velocities. The dotted line is static pressure measured with a piccolo.

Why is the difference important? It was established way back (see the NACA papers) that the front of a cylinder (or top in our case) with the usual half-baffle is mostly cooled by random turbulence. It is static pressure that pushes air down through the baffled fins and out a reduced exit area.

Bottom line? I'd suggest tubes like Section A-A should be used to examine local flows, read individually (example, a hot cylinder problem), not ganged together for an average like in Fig 11-1. As previously noted, I think piccolos are a better choice for investigating upper and lower pressures in RV cowls with various inlets.

Bobby, I sure hope you'll make it a point to join us at the Oshkosh RV Social. I'd really enjoy talking with you.
 
Last edited:
Last weekend I noticed these baffles on a Grumman Tiger.

Left front (#2) cylinder:

e71lp4.jpg


Right rear (#3) cylinder, from top:

2lj5emv.jpg


Right rear (#3) cylinder, from rear:

2zj9f2w.jpg
 
Last weekend I noticed these baffles on a Grumman Tiger...

Looks like Grumman recognizes the need to get air to the bottom fins and provided ample flow.

While better than the typical Vans baffles, is still falls short of a comprehensive airflow management plan - The upper fins are still left to their own and are essentially just radiating in nearly still air of the upper plenum.

I have some changes to the Rocket in work which serve to pull air completely through the fins of the head and cylinders using some intercylinder baffles I have fabricated. In theory, forcing incoming air through all the fin area available should result in much more effective cooling. We'll find out shortly if theory and reality agree.
 
Last edited:
I have a cooling problem in my RV-8A. Cylinder #3 runs very hot on T/O - I've seen that temp spike to ~450 deg before transitioning to cruise/level flight where it settles down to about 360 or so. I should explain that I'm the 3rd owner of the aircraft, and have been modifying it over the past 10 months. Part of the mod process was to replace the non/malfunctioning EGT/CHT gauge with the integrated G3X system, and this is when I discovered the hot cylinder.

Mitigation steps thus far:

1. Replaced the stock baffling with continuous silicon strips. When not in flight, you can see the side baffles drooping away from the cowling, while the rear baffles are compressed against it. Presumably when airflow enters the inlets, the side baffles unfurl to seal the gap. The underside of the cowling has telltale staining - not unlike the Shroud of Turin - that mark where the baffles are rubbing/sealing, and clean gaps where they are not.

2. Seal off the inboard cavity of the inlet ramps formed in the cowling. This was suggested by another builder on my field, so I closed offf the side of the ramps closest to the spinner with fiberglass.

However, there are other issues that I probably need to address, based on the pressure differential discussion in this thread.

There are six blast tubes installed. One on the lower inlet ramp plate for the alternator; two for the mags; one for the battery; one for the fuel pump; and one for the gascolator, all coming out of the back baffle plate. Two questions: Are these blast tubes actually effective, and considering that each of these openings might account for -1" of pressure differential, might they be part of my problem?

The heating system has a 2" scat tube that runs from the back baffle plate, to a muff around an exhaust pipe, to the heater box on the fire wall. When the diverter is closed, it dumps the heated air into the rear of the engine compartment. Based on this thread, it seems to me that this arrangement might have a negative effect on pressure differential.

Anyway, comments are appreciated!

CA
 
Last edited:
Given the thousands of RVs with the identical cabin heater scat tube arrangement, I doubt that is a major contributor to your problem.
 
There are six blast tubes installed. One on the lower inlet ramp plate for the alternator; two for the mags; one for the battery; one for the fuel pump; and one for the gascolator, all coming out of the back baffle plate. Two questions: Are these blast tubes actually effective, and considering that each of these openings might account for -1" of pressure differential, might they be part of my problem?

Charley, here we have a perfect example of where a few measurements would be worth 1000 opinions.

I would not hesitate to fly with those blast tubes blocked (my own airplane has none), so there is an opportunity to obtain "with and without" temperature values if you wish. Likewise, upper and lower plenum pressures, with and without the blast tubes. Send an email if you want to do it and I'll forward a how-to paper.
 
A blast tube is only effective if it it's outlet is very close to the object you are trying to cool, or it is routed to some type of shield/shroud that surrounds it.

Plane Power engineering has said their Alt does not require cooling... that they have tested to well above the temp that the alt gets exposed to where it is located.

The batteries we use have done fine for many years with no indications that the heat they are exposed to is a problem.

A few people have found that cooling the gascolator has helped with a vapor lock problem, but my opinion is that the vapor lock issue is usually caused by something else (I.E., a hot gascolator is not the root problem). The majority of RV's flying with a carb. and gascolator have no gascolator cooling, and don't have vapor lock problems.

Do a simple test.
Block off all of your blast tube inlets with tape and see if it makes a noticeable difference. If not, then you know that you have other issues to resolve (which you may have anyway, but all of the blast tubes could be amplifying the problem).

Edit: Looks like Dan types faster than I do (not hard).
 
Back
Top