What's new
Van's Air Force

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

Test stand detonation data??

Carl Froehlich

Well Known Member
Does anyone have data on conditions to cause detonation in a parallel valve IO-360 or angle head IO-360/390?

I’ll soon need to tell Thunderbolt what pistons to put into my new RV-10 IO-540. I’d like to have 9 to 1 to gain cruise efficiency, but assume the new 100UL will be too pricey to be limited to that one fuel. Looking options to run it on either 93UL or Swift 94UL by:
- Procedurally restricting takeoff manifold pressure
- Retarding timing
- Combination of the two

So far I have information on what people “think” is the way to do this. I’m hopeful this data will provide a firmer foundation.

Carl
 
Sory, no data here. I would expect most test data from the dyno would be at std timing advance. Even more doubtfull to find that data on swift fuel. Most detonation prone environmnt would be high CHTs, high IAT and high MAP. I suspect that retarding timing 5* at high MAP levels would be more than adequate to offset the extra .5 CR and likely will take very little HP off the table and the PV engines are not really prone to detonation. My ignition is set up for 21* above 28" and testing shows no meaningful power reduction.
 
Last edited:
I am interested in this too Carl. Great question, hopefully some experts more knowledgeable than I will be along.
 
There are a few things to note on most detonation testing. The biggest is the temps they test the engine at. The one study required the following and it’s pretty much what most studies use. On a lycoming that means CHT’s between 450 and 490!

TABLE 1. PARAMETER SETTINGS FOR OCTANE RATINGS
Parameter: Maximum Cylinder Head Temperature Within 10°F of maximum recommended by manufacturer All Other Cylinder Head Temperatures Within 50°F of maximum cylinder head temperature Induction Air Temperature Within 4°F of 103°F Induction Air Relative Humidity Less than 30% Oil Temperature Within 10°F of engine manufacturer's recommended maximum.

I looked extensively at the possibility of running auto fuel in my engine and what changes I would need to do it safely. My biggest issue that I have posted here before is the shelf life on auto gas. Octane does not appear to be a big issue with normal engine temps. AvGas is actually 96 octane using the same rating method as auto gas. I also came across this on a article geared toward race engines that compared Race gas, auto gas and AvGas. Might give some pause on running a lot of advance with auto fuel.

Advancing timing on your motor will definitely help with AV Gas and Race Gas due to its slow burn characteristics. On the other hand, be careful if your running commercial grade gasoline, more timing can cause detonation/preignition quite quickly.
 
Last edited:
Does anyone have data on conditions to cause detonation in a parallel valve IO-360 or angle head IO-360/390?

Parallel valve data is scarce. There is some angle valve data in the wild, but it won't be valid for a 9:1 parallel valve 540. Sky Dynamics and Lycon have probably explored parallel valve CR vs detonation limits, but likely did so for 100LL only. And the private player who has extensively explored octane vs detonation (George Braly) likely did so with the typical certification engines, like a 540K and a turbo Continental.

I am endlessly amazed with the RV community's fixation on power. More is good for climb performance, but we already enjoy a high power-to-weight ratio, and roughly half the fleet can't cool stock power anyway. For sure it's a crappy path to cruise efficiency. Rather than trying to keep the pin in a high compression hand grenade, set up a mild motor and work on cooling and drag reduction. It makes everything better, and has no operating cost.
 
However, short of turbonormalizing, a bump in CR does provide instant gains in combustion efficiency at altitude. I considered this route for the Rocket for exactly the same reason. I stuck with 8.5 pistons and clearly I have plenty of power, but the short wings and "low" compression does show up above 9,500 feet.
 
However, short of turbonormalizing, a bump in CR does provide instant gains in combustion efficiency at altitude. I considered this route for the Rocket for exactly the same reason. I stuck with 8.5 pistons and clearly I have plenty of power, but the short wings and "low" compression does show up above 9,500 feet.

Mike, I agree. I'm suggesting the better path would be longer wings. Perhaps not literally, but for sure, better aerodynamics.
 
However, short of turbonormalizing, a bump in CR does provide instant gains in combustion efficiency at altitude. I considered this route for the Rocket for exactly the same reason. I stuck with 8.5 pistons and clearly I have plenty of power, but the short wings and "low" compression does show up above 9,500 feet.

The exact reason for my wanting to go up to 9-1 pistons. I’m not looking for a power bump to enter any “time to climb” competition.

My neighbor Bill Harrelson (Lancair IV that set various world records, including non-stop Guam to Jacksonville FL) has his IO-550 non turbo engine with 10-1 piston for cruise effecincy. Other than the few times he was taking off with ~380 gallons of fuel, he limits takeoff MP to ~25”. He does has EI with ~9 degrees of advance. He just rolled 2100 hours on this engine - the same engine that few get 1500 hours out of before overhaul.

But - he is 100LL (or 100UL) locked.

So I’m back to “we think” ideas on retarding timing and such but with no data to ease my concerns that this is a viable approach.

I contacted Lycoming on this - and got the answer I expected. Use the recommended fuel.
Carl
 
Mike, I agree. I'm suggesting the better path would be longer wings. Perhaps not literally, but for sure, better aerodynamics.

Agree. Longer wings, more fuel, no fuel stops. My Comanche has plenty of draggy things on it but it can easily outrun a RV, more so the higher I go, and have 86 usable gallons. I recently flew a Comanche 400 back to Indiana from Reno NV and average GS was around 195kts, ~17GPH at 15K. One leg was a bit over 1000NM, 5.1 hours non-stop, and had 44 gallons left over. No RV can do anything close to that. Of course 400hp helps :). But wingspan is key. 400 carries 130 gal. of fuel.
 
Agree. Longer wings, more fuel, no fuel stops. My Comanche has plenty of draggy things on it but it can easily outrun a RV, more so the higher I go, and have 86 usable gallons. I recently flew a Comanche 400 back to Indiana from Reno NV and average GS was around 195kts, ~17GPH at 15K. One leg was a bit over 1000NM, 5.1 hours non-stop, and had 44 gallons left over. No RV can do anything close to that. Of course 400hp helps :). But wingspan is key. 400 carries 130 gal. of fuel.

I have an RV-10 with 290 hp and 75 gallons of gas and I see 172kts all the time at 10k and 11 gph. That’s about 40 min difference and 24 gallons less…and I can do my own maintenance. RV-10 for the win!
 
So I’m back to “we think” ideas on retarding timing and such but with no data to ease my concerns that this is a viable approach.

So quit worrying about how to run a hot rod on whale oil. Go mild motor and better aero.

Snip from a Kitplanes piece, re the latest RV-14 changes:

The new configuration incorporates improvements to the cowl’s external shape and cooling path. To illustrate the effect of just one of those changes, tests with the variable exit area (aka the “cowl flap”) fully open, then fully closed, without changing power (approximately 75%) or altitude were 197 and 200 mph, respectively. Repeat the drag calculations, and we find the variation in exit area alone, with its corresponding exit velocity increase and mass flow decrease, changes drag by 4.4%. Need a real-eye-opener? Consider how much power would be required to gain the same 3 mph without a drag reduction:

215 x (200 / 197)³ = 225 HP
225 – 215 = 10 HP

 
The exact reason for my wanting to go up to 9-1 pistons....

So I’m back to “we think” ideas on retarding timing and such but with no data to ease my concerns that this is a viable approach....
Carl

Intuitively, pulling a bunch of timing out increases detonation margin, but how much, who knows? I have data to show that you dont lose much TO power with a big retard, but thats all I was investigating. I can tell you that I have fed the Rocket with 100% auto fuel (cheap, California, regular, at that) in the middle of a Mojave summer and did not get detonation with the retard programmed in. I can also tell you that my neighbor has the SDS ignition and retained his data plate timing for TO power. His is a 10-1 540 and would get detonation even with 100LL. He then switched to "my" curve and the detonation events disappeared completely. Of course CHT and oil temps came down too.

I am very happy with my decision to run piston squirters too. A very simple add on to a 540.
 
Does anyone have data on conditions to cause detonation in a parallel valve IO-360 or angle head IO-360/390?

I’ll soon need to tell Thunderbolt what pistons to put into my new RV-10 IO-540. I’d like to have 9 to 1 to gain cruise efficiency, but assume the new 100UL will be too pricey to be limited to that one fuel. Looking options to run it on either 93UL or Swift 94UL by:
- Procedurally restricting takeoff manifold pressure
- Retarding timing
- Combination of the two

So far I have information on what people “think” is the way to do this. I’m hopeful this data will provide a firmer foundation.

Carl

I did a first run and dyno test my IO-540 at LyCon in December, which has 9:1 Combustion Tech pistons, LyCon ported heads, Sky Dynamics intake and SDS EFII. We did most of the testing on 100LL, but did run some 91 MOGAS and found it ran very well at 25º advance. We also ran at 24º to simulate a hot summer day taking off at full power on 91AKI fuel. It ran very well with a minor reduction in power:

At 2700RPM, 30.2 MP and Vetterman exhaust, 66º OAT at 100º ROP.
319HP at 23º 91 AKI
325HP at 24º 91 AKI
336HP at 25º 100LL
346HP at 28º 100LL

There were absolutely no indications of detonation and a borescope of the cylinders showed the expected "new conditions" of the cylinders.

I can say you should have no issue with 91 AKI fuel at 24º advance. You can run much more advance at lower MAP settings (i.e. at altitude).

Depending on your engine setup, you should have no issue meeting required power for the RV-10 with 91AKI fuel and 9:1 pistons and would expect no issue at all with 94UL, which is a MON rating and has a much larger margin to detonation compared to 91 AKI (MOGAS) fuel.
 
Last edited:
I don’t think so. More like 104. It goes the other way.

This quote might be incorrect but here is where I read that.

Key points:
-100LL is not 100 octane as we rate it for pump gas (R+M/2). It is actually ~96 octane by that measure.
-Quality and consistency is better than pump
-Shelf life is better than pump.
 
I did a first run and dyno test my IO-540 at LyCon in December, which has 9:1 Combustion Tech pistons, LyCon ported heads, Sky Dynamics intake and SDS EFII. We did most of the testing on 100LL, but did run some 91 MOGAS and found it ran very well at 25º advance. We also ran at 24º to simulate a hot summer day taking off at full power on 91AKI fuel. It ran very well with a minor reduction in power:

At 2700RPM, 30.2 MP and Vetterman exhaust, 66º OAT at 100º ROP.
319HP at 23º 91 AKI
325HP at 24º 91 AKI
336HP at 25% 100LL
346HP at 28% 100LL

Ron, that's good data, and pretty much as expected for an parallel valve. I've seen similar timing vs power numbers from the Sky Dynamics dyno.

However, regarding detonation, what was CHT and intake air temperature? As I recall, Lycon's test stand is the propeller type, no cooling air control or intake temperature control.

The detonation test standard requires the cylinder to be at max allowable CHT, all others to be with 50F of max, and IAT to be 100F. Under these conditions, a standard angle valve cylinder run at 20 BTDC can be made to detonate on 100LL. Example attached. Just put the knobs in the wrong place.

If the PIC is a smart engine manager, he/she won't go where there be dragons, but it's not design assurance.
.
 

Attachments

  • LYC 28.6 2400 Detonation.jpg
    LYC 28.6 2400 Detonation.jpg
    139.5 KB · Views: 128
  • LYC Detonation Survey 360 WOT-2700.jpg
    LYC Detonation Survey 360 WOT-2700.jpg
    140 KB · Views: 107
FWIW, I've got an IO-375 on order from Aero Sport with a 9:0 pistons. In discussion with them, they said they're looking at ways to aleviate the requirement for 100LL with those pistons. They may have information you can use.
 
Ron - great data feed, thanks. This is the stuff I’m looking for.

If you have the test details that Dan discussed I’m sure we would like to know.

Carl
 
...

However, regarding detonation, what was CHT and intake air temperature? As I recall, Lycon's test stand is the propeller type, no cooling air control or intake temperature control.

The detonation test standard requires the cylinder to be at max allowable CHT, all others to be with 50F of max, and IAT to be 100F. ...

.

Furthermore that's stated in the research document link I posted considerably earlier which seems like no one has bothered to read?

Also notable in that document - some engines need the most octane for take-off (2700 rpm full throttle, full rich) - and other need most octane running LOP! Not what one would think and why the document tests 7 different regimes.
 
Ron, that's good data, and pretty much as expected for an parallel valve. I've seen similar timing vs power numbers from the Sky Dynamics dyno.

However, regarding detonation, what was CHT and intake air temperature? As I recall, Lycon's test stand is the propeller type, no cooling air control or intake temperature control.

.

Dan, that is correct, it is a propeller type test stand with no intake temperature control. Cylinder cooling is simply propeller air over the unbaffled cylinders. Outside Air Temperature varied from 60-68º during the days there, however intake air temperature was 66ºF for test runs I mentioned, as measured at the throttle body (one probe for each ECU).

CHT's went above 380º in all cases and test runs were stopped (reduced power) once one CHT reached 420º (almost always cylinder #6 for all tests). LyCon's testing protocols limit CHT to 420º, and if we're being honest, you shouldn't let your cylinders go above 400º during normal operation if you want to have extended cylinder life.

Yes, Lycoming testing protocol tests at 500º: my testing was not intended to be destructive, but rather informative for real-world operation of my engine. And I'd assume that is what most folks want to know if they're considering an engine setup for their airplane.

Having the ability to trim fuel to each cylinder allowed getting all cylinders within 30º of each other, so one number is fairly indicative of the engine overall. I also expect that once the engine is baffled, and fuel trim is adjusted further, ability to maintain CHT's well below 400º will be greatly improved and more evenly matched.

One somewhat unrelated data point, more of a plug for the SkyDynamics intake, is that we saw a rise in manifold pressure from the throttle body MAP to the cylinder intake ports of just over 1.2" at 2700 RPM, fairly consistent across all cylinders.
 

Attachments

  • A45F5B8D-E8E7-48F3-91B7-78BF91133669.jpg
    A45F5B8D-E8E7-48F3-91B7-78BF91133669.jpg
    502.9 KB · Views: 144
Last edited:
Furthermore that's stated in the research document link I posted considerably earlier which seems like no one has bothered to read?

Bryan,

I very much wanted to read it but to do that I had to register to download the reader - which I did not do. I get enough SPAM already.

Carl
 
Furthermore that's stated in the research document link I posted considerably earlier which seems like no one has bothered to read?

Also notable in that document - some engines need the most octane for take-off (2700 rpm full throttle, full rich) - and other need most octane running LOP! Not what one would think and why the document tests 7 different regimes.

I've read this study before.

This was a massive test however the presentation of data in the charts is truly horrible, almost unreadable without a lot of patience, having entries running through adjacent boxes. They should have graphed results as Lycoming does- much more reader friendly presentation for data comparison.

Interesting and somewhat questionable that they found higher octane requirements at lower rpm and MAP when running lean on the 320. BMEP MUST be lower there, although EFFECTIVE timing as related to rpm is higher as rpm is decreased (lower piston speed).

The AFR data is buried later in the document however it's difficult to cross reference this with other variable changes. Would have been easier to state these in the same charts and also reference the number of degrees ROP or LOP plus the AFR data.

Other dyno data I've seen on PV engines suggests that propensity for detonation is worst somewhat ROP (say 50F). When you lean from a overly(and possibly unknown) rich position, you may cross this zone and not in fact be LOP. This is important data to know and may lead to erroneous conclusions.

Both Mike Robinson and Greg Niehues run mogas and have a lot of hours in hot climates on PV engines. They would be good sources for info on safe limits for MAP, mixture, RPM, IAT, CHT and timing as they've accumulated a lot of flight time with no apparent detonation issues.
 
Last edited:
Furthermore that's stated in the research document link I posted considerably earlier which seems like no one has bothered to read?

Also notable in that document - some engines need the most octane for take-off (2700 rpm full throttle, full rich) - and other need most octane running LOP! Not what one would think and why the document tests 7 different regimes.

This would be contrary to typical premixed flame behavior (relevant here) but that's why we test. Years back my employer "asked" me to take a grad level course on combustion. My head still hurts recalling it. The notes and textbook are still in the office but will try and add a graphic here tomorrow. For gas turbines to meet emissions (NOx) requirements, the flame overall temp is 2200 degF-ish. Lower would be better and there's some margin before CO emissions start to rise but we're already dancing close to the edge of theoretic combustion limits, = flame out. While the types of combustion, detonation (sonic) and deflagration (subsonic) can be mixed on either side of stoichiometry, the trends are very clear. Rich trends toward strong detonation and vice versa. Not saying the data is wrong but it is quite surprising. Again, this is why we test.

The test stand photo shown is NOT a good representation of engine operation as it appears there's no plenum to direct down-flow through cylinders. It could be but I'm not convinced it is a conservative approach. It's very possible, probably guaranteed, that some cylinder hot spots exists. This can lead to pre-ignition that can skew the data if not detected and accounted for.
 
Last edited:
The test stand photo shown is NOT a good representation of engine operation as it appears there's no plenum to direct down-flow through cylinders. It could be but I'm not convinced it is a conservative approach. It's very possible, probably guaranteed, that some cylinder hot spots exists. This can lead to pre-ignition that can skew the data if not detected and accounted for.

There is in-fact no plenum or baffling for cylinder airflow on LyCon's engine test stand. They've been testing engines this way for several decades; works well for it's intended purpose. A normally baffled engine would have better cylinder cooling.
 
Ron, thanks for posting info from your dyno runs at Ly-Con. Data is almost always better than conjecture and I think others here will find it interesting and useful.
 
This quote might be incorrect but here is where I read that.

Key points:
-100LL is not 100 octane as we rate it for pump gas (R+M/2). It is actually ~96 octane by that measure.
-Quality and consistency is better than pump
-Shelf life is better than pump.

87 auto gas is about 84 motor octane (‘M’) and 90 research (‘R’) octane, average (R+M)/2=87. 100LL av gas is motor octane (‘M’) only. Research octane would be about 108.
 
Can't add an attachment in an edit. Now I know.

Attached is the graphic mentioned in my previous post from my ancient class notes (yes, I have a problem) as I couldn't find the text book. Detonation can happen at any AF ratio but it usually takes some other influence like a hot spot (surface ignition), flash-back (continuous flow system) unstable mixture/combustion, etc when on the lean side. I'll state again the LOP detonation in a recip engine is a big time surprise from my understanding of the related physics. Don't want to start a fight but I wouldn't be surprised if there was some test set-up discrepancy or possibly some strong combustion dynamics misinterpreted as pinging. Dynamics are very common with very low AF ratios. Cool finding for sure. I work with a bunch of talented combustion engineers. I'll see if they have any ideas. Their work is gas turbine combustor related so I'm not too hopeful. Wickedly intelligent people but the sky tends to be a different color in their world.
 

Attachments

  • 14MAR2022.pdf
    373.2 KB · Views: 231
. I'll state again the LOP detonation in a recip engine is a big time surprise from my understanding of the related physics. Don't want to start a fight but I wouldn't be surprised if there was some test set-up discrepancy or possibly some strong combustion dynamics misinterpreted as pinging.

I agree. The scenario they found in this test seems questionable. PCP and flame speed are both lower LOP than at ROP and these are the major detonation drivers involved if IAT and CHT remain similar. In fact LOP also reduces CHT, giving further detonation margin.
 
Bryan,

I very much wanted to read it but to do that I had to register to download the reader - which I did not do. I get enough SPAM already.

Carl

I just checked the link and I didn't have to login or register. Clicked the Download button, checked the 'I am not a robot', and then it made me wait 30 seconds and gave me a button to download it. If you are still stuck, PM me your email and I'll forward it to you directly.


Regarding readability... well it is a technical document written by engineers, with a lot of consideration of how to do the test, how they did the test, etc. etc. At least you shouldn't have outstanding questions about what they did when you get through it.
 
We must not be reading the same document. There is no lean of peak operation in AR 99/70. They leaned from full rich in 5% increments until knock was detected or they reached peak EGT.

Two snips from the report:

If knock is not detected while operating on the specific reference fuel, a 5% lean condition is then tested. Leaning is performed from the full-rich reading at the cruise power setting. Conditions are allowed to stabilize after adjusting the mixture. Knock data are then recorded. If knock is detected, the house fuel is selected and the cruise setting with a 5% lean mixture setting is retested using a higher octane reference fuel. If knock is not detected, the mixture is leaned by the 5% increment previously determined. This 5% increment leaning is continued until either knock is detected or peak EGT is eclipsed.

It is important to note that all lean conditions represent percentage reductions in fuel flow from the full-rich mixture position while operating on the particular rating fuel. For example, to test the cruise position at 5% lean condition on 98 MON reference fuel, it would require determining the full-rich fuel flow rate while operating on 98 MON reference fuel at the cruise position and then adjusting the mixture to reduce the fuel flow rate by 5%.
 
We must not be reading the same document. There is no lean of peak operation in AR 99/70. They leaned from full rich in 5% increments until knock was detected or they reached peak EGT.

Two snips from the report:

If knock is not detected while operating on the specific reference fuel, a 5% lean condition is then tested. Leaning is performed from the full-rich reading at the cruise power setting. Conditions are allowed to stabilize after adjusting the mixture. Knock data are then recorded. If knock is detected, the house fuel is selected and the cruise setting with a 5% lean mixture setting is retested using a higher octane reference fuel. If knock is not detected, the mixture is leaned by the 5% increment previously determined. This 5% increment leaning is continued until either knock is detected or peak EGT is eclipsed.

It is important to note that all lean conditions represent percentage reductions in fuel flow from the full-rich mixture position while operating on the particular rating fuel. For example, to test the cruise position at 5% lean condition on 98 MON reference fuel, it would require determining the full-rich fuel flow rate while operating on 98 MON reference fuel at the cruise position and then adjusting the mixture to reduce the fuel flow rate by 5%.

Ah, my mistake! I misread "15% Lean" to mean LOP. It is really "15% lean of full rich"... which I guess is anyone's guess as to whether that is 50 degrees ROP or wherever. They did note that "10% lean point fell slightly lean of best power." This is all under table 4. While table 4 is for the Continental IO-550-D I think the important take away is that the IO-550-D behaves considerably differently under leaner conditions that the Lycoming 320-B did. So trying to use results on a single engine such as the 320-B and extrapolating that to all Lycomings... seems risky.
 
Yes, Lycoming testing protocol tests at 500º: my testing was not intended to be destructive, but rather informative for real-world operation of my engine. And I'd assume that is what most folks want to know if they're considering an engine setup for their airplane.

Absolutely, but with a caveat. Fundamentally, those are best power runs. We already know staying well on the rich side of peak is fairly safe, 100 ROP being the arbitrary point we typically accept as the warning line. The torque curve is very flat around best power mixture, so there is no value in testing leaner if power is the primary interest, on any fuel.

Here the OP's professed interest is cruise efficiency, i.e. leaned much further, at less manifold pressure and RPM, and he will no doubt be doing it with whatever ignition timing the E-Mag folks are supplying with the 6-cyl p-mag.

So, yes, I too love the data, but....

One somewhat unrelated data point, more of a plug for the SkyDynamics intake, is that we saw a rise in manifold pressure from the throttle body MAP to the cylinder intake ports of just over 1.2" at 2700 RPM, fairly consistent across all cylinders.

Can I interest you in connecting a delta-p sensor to a primer port? I have some 720 degree port pressure data for a standard Lycoming horizontal plenum sump. I'm guessing the above is an average. The delta-p sensor and a laptop capture yields wave amplitudes, with timing.
 
SNIP

Here the OP's professed interest is cruise efficiency, i.e. leaned much further, at less manifold pressure and RPM, and he will no doubt be doing it with whatever ignition timing the E-Mag folks are supplying with the 6-cyl p-mag. SNIP

Dan is correct, I’d like to operate a Cold Air Sump IO-540 with 9-1 pistons on Swift 94 and/or 93 ethanol free mogas. The stock 8.5-1 pistons support this but I was hoping for some data on to implement operational mitigations to run the higher compression pistons (such as retarded timing and/or operational MP pressure limitations on takeoff).

Side note - having installed the six cylinder pMag on my old RV-10 base timing and maximum advance is set on installation. Typically I consider 8-9 degrees for the maximum advance setting. Also note this pMag has switchable timing (while the engine is running) to lock in timing to whatever you set for base timing - typically 25 degrees). Here I see this as another means to mitgate risk as, for example, setting base timing at 20 degrees for take off and climb with the mode switch in “fixed”, then moving the mode switch to “variable” after setting up for LOP cruise.

Carl
 
Also note this pMag has switchable timing (while the engine is running) to lock in timing to whatever you set for base timing - typically 25 degrees). Here I see this as another means to mitgate risk as, for example, setting base timing at 20 degrees for take off and climb with the mode switch in “fixed”, then moving the mode switch to “variable” after setting up for LOP cruise.

That's a good feature for a 540 parallel valve. If you can also set your own advance schedule, I'd quit worrying. You'll be able to run your own tests, flying progressively more advanced schedules, with the option to revert to fixed timing if/when you see a abnormal CHT rise and a dropping EGT.

Go with fixed for ROP climb too.
 
Last edited:
That's a good feature for a 540 parallel valve. If you can also set your own advance schedule, I'd quit worrying. You'll be able to run your own tests, flying progressively more advanced advance schedules, with the option to revert to fixed timing if/when you see a abnormal CHT rise and a dropping EGT.

Go with fixed for ROP climb too.

Correct, you can set the advance from O degrees to something around 15 if memory serves (but again anything more than 8 or 9 degrees is not what normal RVs want to do).

Carl
 
Dan is correct, I’d like to operate a Cold Air Sump IO-540 with 9-1 pistons on Swift 94 and/or 93 ethanol free mogas. The stock 8.5-1 pistons support this but I was hoping for some data on to implement operational mitigations to run the higher compression pistons (such as retarded timing and/or operational MP pressure limitations on takeoff).

Carl, I provided data earlier showing that you can run 9:1 pistons on 91AKI fuel without issue. 94UL Swift fuel, or G100UL, both of which claim to be more widely distributed in the next year, provide a much higher margin to detonation than 91AKI pump gas.

Normal timing for an IO-540 is 25º fixed advance. While you can retard much more if desired, you likely won't need to go much more than 23-24º to provide a safe increase in margin to detonation - something I would only consider if OAT was over 100ºF and needed to take off at full throttle (and yes, an RV-10 will take off just fine at 2500 RPM as well).

I have no concerns taking off with MOGAS on 9:1 pistons; however, you cannot run 91AKI fuel at the worst case scenario conditions of the Lycoming test protocol. It will require you properly manage your mixture and be mindful of CHT peak temperatures.

If you are experiencing, or expect to experience, CHT's exceeding 420º during climb out, then you might consider 8.5:1 pistons to provide you the extra margin to detonation, but I'd also encourage you to consider ensuring good cooling to your cylinders and attempt to limit CHT to less than 400º for extended cylinder life.

Can I interest you in connecting a delta-p sensor to a primer port? I have some 720 degree port pressure data for a standard Lycoming horizontal plenum sump. I'm guessing the above is an average. The delta-p sensor and a laptop capture yields wave amplitudes, with timing.

I'm always interested in a fun experiment... At the moment, I have my engine in lay-up pending finishing of the airplane (airport move someday soon...Tuesday?). I'd love to see how the intake responds with changes in timing - something I hadn't considered.

Just to be certain, the pressures I reported are differential between sensors at the throttle body compared to a pressure sensor I moved to each of the six cylinder priming ports at full power.
 
Last edited:
Hi Ron, I am not 100% on this but I am sure the smarter people here will correct me if wrong but if you have a concern for detonation on takeoff I would run 2700 not 2500 RPM. Like a car high manifold pressure combined with low rpm decreases your detonation margin. Best case would be to run 2700 RPM but reduce manifold pressure for takeoff but now you get into other issues especially if carbed regarding fuel flow. Here is a bit from Pelicans perch.

“Take RPM, for example. By reducing RPM, you slow the engine, so the crankshaft is turning more slowly. The combustion event still takes roughly the same time to finish (give me a little room here) but the crank hasn’t turned as far. Result, the peak pressure and temperature occur closer to TDC, decreasing the detonation margin.”
 
Carl, I provided data earlier showing that you can run 9:1 pistons on 91AKI fuel without issue....

The detonation-free runs at Lycon were at 100 ROP. A standard test sweeps mixture through the entire range. Consider this question...what happens if you develop a partially plugged nozzle and go to full throttle? You're not likely to notice during runup.

I'd love to see how the intake responds with changes in timing...

I'm sorry, I wasn't clear. When I said "timing", I meant when pressure waves arrive at the intake port. The desired wave timing is a strong positive at intake opening, and another just prior to close.
 
Hi Ron, I am not 100% on this but I am sure the smarter people here will correct me if wrong but if you have a concern for detonation on takeoff I would run 2700 not 2500 RPM. Like a car high manifold pressure combined with low rpm decreases your detonation margin. Best case would be to run 2700 RPM but reduce manifold pressure for takeoff but now you get into other issues especially if carbed regarding fuel flow. Here is a bit from Pelicans perch.

“Take RPM, for example. By reducing RPM, you slow the engine, so the crankshaft is turning more slowly. The combustion event still takes roughly the same time to finish (give me a little room here) but the crank hasn’t turned as far. Result, the peak pressure and temperature occur closer to TDC, decreasing the detonation margin.”

RPM has an impact on where optimum timing lies, but RPM by itself has no meaningfull bearing on detonation. The most detonation prone environments are heavy load or lugging. For example in your manual trans car, going to WOT without downshifting is the most typical and likely scenario for pinging. Do the same thing, but with a downshift first and there is no pinging. Increased RPM is your friend here eventhough the RPM itself isn't reducing the det potential, it is the reduced load at the higher gear ratio. The side effect is higher RPM.


I would argue that on a CS prop airplane, 2700 and WOT is significantly less detonation prone than 2500 and WOT for the reason described above. The higher RPM setting will cause a finer blade setting, producing less load on the engine and therefore less likely to detonate. The Pilot's method for reducing detonation potential on TO is 2700 RPM and limiting MAP to something less than 30". Reducing timing and keeping CHT and IAT at reasonable levels is a better way, but typically not in the pilots control.

Avoiding detonation in the climb would be to keep RPM at 2700 and reduce MAP. This is better than pulling the RPM back to 2500. Again, reducing the load on the engine. With the same HP / thrust production, The 2700 RPM setting will be a lower load on the engine than 2500.

Larry
 
Last edited:
@lr172

Great stuff Larry, as always.

Could you educate us on the detonation relationships with displacement, specifically bore. It seems intuitive based on relative FA charge amounts (energy) that smaller equates to more detonation margin. This is anecdotally supported by the plethora of high revving, small bore engines on motorcycles surviving just fine very high with CRs; 15-1, 16-1, etc. Can you make us (i.e. me) smarter here?
 
Not sure where I read this, but for a gasoline engine the bore limit (for aircraft engines at least) seems to be 6". Thus more and more cylinders, RR Merlin and Allison V-12s, two row radials of 2800 cu in, 4 row radials of 4360 cu in. I don't think any of these bores exceed 6".
 
Practical Applications

Guys, while I truly enjoy theoretical discussions, we're not on the same page regarding the OP's original question. Rather than create more uncertainty and doubt, I'd prefer we look for ways to provide practical - measurable and controllable, methods and parameters for the average pilot to work with.

While detonation is a real thing, we don't have a "margin to detonation" meter. The number one parameter that should be measured is cylinder head temperature. The number one way for a pilot to control that is with the throttle (manifold pressure) and then with mixture (fuel flow). I would hope we can agree that lower power and cooler cylinders provide more margin from detonation.

For my engine at full throttle (same manifold pressure set point), I provided measured data points, not theoretical. And when you discuss "load", I assume you mean torque, because that is what the crankshaft is seeing. Torque is what is measured - horsepower is calculated.

The devil's in the details; you can come up with many different settings to achieve your goals with engine power, but my point is simply a response to the main concern for this thread, that I understood was, "I'd like to run 9:1 pistons on 93UL, or Swift 94, gas". The answer to that is yes, you can do that.
 
Last edited:
For my engine at full throttle (same manifold pressure set point), 2700 RPM will provide about 670 ft-lbs of torque to the prop. 2500 RPM will provide about 550 ft-lbs of torque. Those are measured data points and not theoretical. That's 120 fl-lbs of torque less at 2500 compared to 2700; torque is what is measured - horsepower is calculated.

Where did you get this data? The lyc torque is pretty flat. I would be hugely shocked if the Lyc lost 20% of it's torque going from 2500 to 2700. If it lost that much, you would see the HP drop as well, as 200 RPM is not a lot of HP adder. I would expect peak torque around 2500 with a very small fall off at 2700.

ok, lets take my 540C4B5 for example. 250 HP at 2575 and 260 HP at 2700 - straight from the Lyc specs. Cam lobe shape and opening / closing angles (what determines the torque curve shape and RPM points on that curve) are essentially the same for the 4 cyl models.

250 x 5252 / 2575 = 509 ft/lbs @ 2575 RPM
260 x 5252 / 2700 = 506 ft/lbs @ 2700 RPM

These are also data points and apparently a lot more reliable than your data points. No theory here. Basic combustion engineering math applied to Lyc published data.

Larry
 
Last edited:
@lr172

Great stuff Larry, as always.

Could you educate us on the detonation relationships with displacement, specifically bore. It seems intuitive based on relative FA charge amounts (energy) that smaller equates to more detonation margin. This is anecdotally supported by the plethora of high revving, small bore engines on motorcycles surviving just fine very high with CRs; 15-1, 16-1, etc. Can you make us (i.e. me) smarter here?

Not an expert on combustion physics, but know that compressed volume shape and special shape features, like a quench area, impact detonation potential the most. Of course, beyond all the other factors mentioned in this thread. Swirling or stagnation of the fuel air mixture have a significant influence on detonation potential, due to the movement of the flame front after ignition. They also influence where best power advance will be for any given set of conditions. I am not aware how or if displacement alone impacts this. My understanding is that is does not.

Larry
 
Last edited:
Where did you get this data? The lyc torque is pretty flat. I would be hugely shocked if the Lyc lost 20% of it's torque going from 2500 to 2700. If it lost that much, you would see the HP drop as well, as 200 RPM is not a lot of HP adder. I would expect peak torque around 2500 with a very small fall off at 2700.

ok, lets take my 540C4B5 for example. 250 HP at 2575 and 260 HP at 2700 - straight from the Lyc specs. Cam profile and opening closing angles (what determines the torque curve shape and RPM points) are essntially the same for the 4 cyl models.

250 x 5252 / 2575 = 509 ft/lbs @ 2575 RPM
260 x 5252 / 2700 = 506 ft/lbs @ 2700 RPM

These are also data points and apparently a lot more reliable than your data points. No theory here. Basic combustion engineering math applied to Lyc published data.

Larry
Please see my previous posts showing my dyno testing at LyCon, which includes both power and torque curves. I personally observed their unit be calibrated, so your comments about my data not being reliable are without merit.

It really comes down to the engine and a tremendous variance in setup and run parameters.
 
Last edited:
Please see my previous posts showing my dyno testing at LyCon. I personally observed their unit be calibrated, so your comments about my data not being reliable are without merit.

Your numbers here:

670 * 2700 / 5252 = 344 HP at 2700
550 * 2500 / 5252 = 260 HP at 2500

Sorry, but this data is GROSSLY inconsistent with Lyc engine performance data seen by the rest of us. Not to mention the fact that I question whether a naturally aspirated PV 540 can produce 344 HP at 2700 with 9:1 pistons.

Larry
 
Last edited:
Back
Top