Matronics Email Lists :: View topic

Matronics Email Lists
Web Forum Interface to the Matronics Email Lists

Get Email Distribution Too! FAQ Search Memberlist Usergroups Register

Profile Log in to check your private messages Log in

Knock sensors?
Goto page 1, 2 Next

Matronics Email Lists Forum Index -> Engines-List

View previous topic :: View next topic

Author Message

Mark Phillips in TN

Joined: 10 Jan 2006
Posts: 431
Location: Columbia, TN

Posted: Thu Oct 11, 2007 7:11 pm Post subject: Knock sensors?

We've had knock sensors on cars for many years that retard timing in the event of detonation to keep engines running at peak fuel efficiency.  Since it is almost impossible to detect thru conventional aircraft engine instrumentation that I'm aware of, or audibly, (such as in cars) that such a device would be very useful in planes, but I've never heard of such a critter.  I'm kinda guessing that it is difficult in an air-cooled engine or because of the large cylinder displacement.  

  

 Anyone aware of any attempts to do this on Lycs or Conts?

  

 Mark Phillips



See what's new [quote][b]

_________________
From The PossumWorks...

GrummanDude

Joined: 15 Jan 2006
Posts: 926
Location: Auburn, CA

Posted: Thu Oct 11, 2007 11:06 pm Post subject: Knock sensors?

I read years ago that knock sensors were tried but didn't work because 

with the piston slap that is present in air cooled aircraft engines 

makes it difficult for the sensor to determine the difference between 

piston slap and 'knock.'

--

_________________
Gary
AuCountry Aviation
Home of Team Grumman

tompkinsl(at)integra.net
Guest

Posted: Fri Oct 12, 2007 11:05 am Post subject: Knock sensors?

Some time ago I spoke to a long-time engineering  friend who has spent his entire (40-yr.) career in automotive engine  development.  My understanding? is that automobile engine knock  sensors  are block mounted and detect vibration in engines that compared to  aircraft engines run quite smooth.  Our engines are s*** shakers, have  individual cylinders.  His conclusion was that automotive typeknock sensors  would probably NOT work in aircraft engines.  I consider him very  knowledgable as  well as an excellent engineer.

  

 Larry Tompkins

   [quote][b]

ainut(at)hiwaay.net
Guest

Posted: Fri Oct 12, 2007 1:15 pm Post subject: Knock sensors?

Knock sensors are tuned to specific frequency bands  depending upon the engine they're used on.  Any engine with such large jugs  that they have to have 2 plugs per cylinder just to fire all the fuel/air  mixture packs a whale of a wollop when it fires.  This power  ulse wollop is so much larger than the majority of engine knocks that the  use of a knock sensor is fruitless.  It's also why some props will not work  with some engines.

  

 David M.

  

 [quote]   ---

azevedoflyer(at)aol.com
Guest

Posted: Fri Oct 12, 2007 5:26 pm Post subject: Knock sensors?

Maybe we should try again!

 Piston slap is a repeatable, rpm-correlated event. Detonation has a high frequency and fast-decaying "signature". Can be easily tracked and filtered. Never had problems with these two. Aircooled engines complicate things a bit because of the finning which radiate vibration in a very ample spectrum. Despite the apparent complexity, I am sure someone can develop a reliable knock sensor for aircooled engines. What about PORSCHE engines? 

 The problem might reside more in the lack of a large enough  market for such product than anything else.

 

 

   

 

   

 

 --

Jim Baker

Joined: 30 Mar 2006
Posts: 181
Location: Sayre, PA

Posted: Fri Oct 12, 2007 5:44 pm Post subject: Knock sensors?

 	  Quote: 	
		   I read years ago that knock sensors were tried but didn't work because 

 with the piston slap that is present in air cooled aircraft engines 

 makes it difficult for the sensor to determine the difference between 

 piston slap and 'knock.'


	


Can't speak to the whole issue but...



I live a short ways from GAMI at Ada, OK. While over there one day I asked 

about their electronic ignition. They're using an optical sensor (laser pulsed 

into the cylinder via a fiber optic cable situated on an offset spark plug). The 

regular 18 mm plug has the cylinder offset axially from the center and the 

fiber optic is positioned on the other side. The laser detects the onset of 

detonation through some sophisticated algorithims looking at charge density 

and flame front propagation (as they explained it to me). GAMI chose the 

nastiest engine known in terms of detonation sensitivity, a TSIO 540. They 

can virtually feed this beast almost anything combustible and it runs without 

complaint. They even demonstarte changing fuel types during a full power 

test cell run and you'd never know anything had changed except by watching 

the timing and EGT changes on the data screen. Its been a year or so since I 

saw it in operation and don't know their current status for certification testing. 



Jim Baker

580.788.2779

'71 SV,  492TC

Elmore City, OK

altoq(at)cebridge.net
Guest

Posted: Fri Oct 12, 2007 6:36 pm Post subject: Knock sensors?

Porsche air cooled six cylanders have one on each  side mounted on a bar that is bolted to three cylinders. They work good.  

 [quote]   ---

klwerner(at)comcast.net
Guest

Posted: Fri Oct 12, 2007 9:42 pm Post subject: Knock sensors?

Aircooled Porsche's are somewhat  around 1/2 Liter per Cylinder (~3Liter / 6Cyl.), whereas the  360cid Lycoming's cylinder volume's are nearly TRIPLE  that...  Different smoothness for sure...

 [quote]   ---

glcasey

Joined: 01 Jul 2007
Posts: 23

Posted: Sat Oct 13, 2007 4:40 am Post subject: Knock sensors?

I suspect you are right - the market is very, very small.  The knock  

sensor itself usually is tuned to respond to the knock frequency of a  

particular engine and the frequency of a large-bore aircraft engine  

is almost certainly a lot different than a car engine, preventing the  

use of a cheap readily available sensor.  Any knock control system is  

software-intensive, with the algorithm being carefully worked out  

during lots of experimentation (running on the dyno for lots of  

hours).  At high power settings the knock is often undetectable as it  

is drowned out by the other noises (the car engine systems that I was  

familiar with years ago essentially shut off at high load and relied  

on the timing acquired during light load operation).  And then the  

real kicker is that on an aircraft system you want some kind of fault  

tolerance - how do you protect against some fault in the system that  

over-advances the spark?  What do you use to check it?  The  

combination of all these things will, in my opinion, prevent the use  

of a vibration-based knock system in an aircraft engine for some  

time.  The optical system is perhaps a different matter.

Gary Casey

 	  Quote: 	
		  


	


 	  Quote: 	
		   Time: 06:26:19 PM PST US

 Subject: Re: Knock sensors?

 From: azevedoflyer(at)aol.com

  Maybe we should try again!

 Piston slap is a repeatable, rpm-correlated event. Detonation has a  

 high frequency

 and fast-decaying "signature". Can be easily tracked and filtered.  

 Never had

 problems with these two. Aircooled engines complicate things a bit  

 because

 of the finning which radiate vibration in a very ample spectrum.  

 Despite the apparent

 complexity, I am sure someone can develop a reliable knock sensor for

 aircooled engines. What about PORSCHE engines?

 The problem might reside more in the lack of a large enough  market  

 for such product

 than anything else.

_________________
Gary Casey

Mark Phillips in TN

Joined: 10 Jan 2006
Posts: 431
Location: Columbia, TN

Posted: Sun Oct 14, 2007 7:26 pm Post subject: Knock sensors?

In a message dated 10/12/2007 8:27:32 PM Central Daylight Time, azevedoflyer(at)aol.com writes:

  	  Quote: 	
		  Aircooled engines complicate things a bit because of the finning which radiate vibration in a very ample spectrum. 	


 >>>

  

 Based on my meager understanding of knock sensor physics in automotive applications, this was my primary suspicion- how to isolate specific "knock" signatures when you have several hundred fins vibrating in multiple frequencies and harmonics of same.  If Lycosaurs were falling from the sky due to detonation at the hands of efficiency-desparate pilots, I'd believe market and/or regulatory pressure would force development of such a system.  Perhaps when domestic airplane fuel hits European price$ and truly efficient operation becomes a defining market factor, this technology will develop and we'll have a new paradigm: LOD operation vs. LOP, just like in your average run-of-the-mill ten-year-old Nissan Sentra....

  

 With the astonishing capacity of modern digital processing, I can't understand how an on-board processor couldn't sample vibratation sendor data from multiple points, compare that against know "safe" data, then somehow annunciate a potentially unsafe condition and/or make appropriate adjustments to whatever engine-management scenario a pilot may employ in the ultimate interest of maximizing efficiency.

  

 Here's hoping that if observers from the Lycoming/Continental intelligentsia have comments, they'd chime in here, even if covertly.

  

 Appreciate the discussion-

 Mark



See what's new at AOL.com and

_________________
From The PossumWorks...

ainut(at)hiwaay.net
Guest

Posted: Sun Oct 14, 2007 8:56 pm Post subject: Knock sensors?

if efficiency were truly a concern, they'd all convert to fuel  injection.  If that, then the pilot would have no control over anything  except which way to point -- and would not need it.  The computer would  control LOP etc.

  

 David M.

  

  

 [quote]   ---

klehman(at)albedo.net
Guest

Posted: Mon Oct 15, 2007 3:50 am Post subject: Knock sensors?

Surely ion sensing of detonation as Saab has been doing (with electronic 

ignition) is cheaper to develop and more precise. Could be it will soon 

be time to advance into the last century with these engines.

Ken



Fiveonepw(at)aol.com wrote:



 	  Quote: 	
		   In a message dated 10/12/2007 8:27:32 PM Central Daylight Time, 

 azevedoflyer(at)aol.com writes:



     Aircooled engines complicate things a bit because of the finning

     which radiate vibration in a very ample spectrum.



 >>>

  

 Based on my meager understanding of knock sensor physics in automotive 

 applications, this was my primary suspicion- how to isolate specific 

 "knock" signatures when you have several hundred fins vibrating in 

 multiple frequencies and harmonics of same.  If Lycosaurs were falling 

 from the sky due to detonation at the hands of efficiency-desparate 

 pilots, I'd believe market and/or regulatory pressure would force 

 development of such a system.  Perhaps when domestic airplane fuel 

 hits European price$ and truly efficient operation becomes a defining 

 market factor, this technology will develop and we'll have a new 

 paradigm: LOD operation vs. LOP, just like in your average 

 run-of-the-mill ten-year-old Nissan Sentra....

  

 With the astonishing capacity of modern digital processing, I can't 

 understand how an on-board processor couldn't sample vibratation 

 sendor data from multiple points, compare that against know "safe" 

 data, then somehow annunciate a potentially unsafe condition and/or 

 make appropriate adjustments to whatever engine-management scenario a 

 pilot may employ in the ultimate interest of maximizing efficiency.

  

 Here's hoping that if observers from the Lycoming/Continental 

 intelligentsia have comments, they'd chime in here, even if covertly.

  

 Appreciate the discussion-

 Mark

apilot2(at)gmail.com
Guest

Posted: Mon Oct 15, 2007 4:27 am Post subject: Knock sensors?

Wrong. Most aircraft engines being sold today are fuel

injected........mechanical, continuous flow. No computer involved,

pilot controls the mixture. For computer control, you have to develop

a mixture map for each variety of engine. Ain't gonna happen. Not to

mention that TCM and Lycoming oppose LOP operations, so they wouldn't

design for it.



On 10/14/07, David M. <ainut(at)hiwaay.net> wrote:

[quote]



 if efficiency were truly a concern, they'd all convert to fuel injection.

 If that, then the pilot would have no control over anything except which way

 to point -- and would not need it.  The computer would control LOP etc.



 David M.



 ---

Float Flyr

Joined: 19 Jul 2006
Posts: 2704
Location: Campbellton, Newfoundland

Posted: Mon Oct 15, 2007 4:32 am Post subject: Knock sensors?

speaking  only as an outsider there are two things that I see regarding the use of knock  sensors.

  

 First  the engines are basically high volume low compression so they are not overly  prone to knocking.  Sure it can/does occur but not as often as on higher  compression engines.

  

 The  second point is any knocking would be fairly low in amplitude compared with a  higher revving high CR engine.

  

 Even so  I think it is obvious that a knock sensor can be developed to operate  efficiently on these air-cooled engines.  I'll bet they could even install  knock sensors in the crank shafts that way sensors could do double duty as  a form of torque meter as well as knock sensor.

  

 Noel

 [quote]   

   --

_________________
Noel Loveys
Kitfox III-A
Aerocet 1100 Floats

Float Flyr

Joined: 19 Jul 2006
Posts: 2704
Location: Campbellton, Newfoundland

Posted: Mon Oct 15, 2007 5:48 am Post subject: Knock sensors?

I've seen the term LOP used a few times..... But have no idea what it means.



Noel 



[quote] --

_________________
Noel Loveys
Kitfox III-A
Aerocet 1100 Floats

gyoung

Joined: 09 Jan 2006
Posts: 211
Location: Republic of Texas

Posted: Mon Oct 15, 2007 2:06 pm Post subject: Knock sensors?

LOP=lean of peak, ROP=rich of peak



Regards,

Greg Young

 



[quote] --

Mark Phillips in TN

Joined: 10 Jan 2006
Posts: 431
Location: Columbia, TN

Posted: Mon Oct 15, 2007 3:20 pm Post subject: Knock sensors?

In a message dated 10/15/2007 8:52:18 AM Central Daylight Time, noelloveys(at)yahoo.ca writes:

  	  Quote: 	
		  I've seen the term LOP used a few times..... But have no idea what it means. 	


 >>>

 Normally for injected engines- stands for Lean Of Peak.  If using an EGT gauge, leaning the mixture will cause EGTs to rise to a maximum, then begin to fall until there is too little fuel to sustain combustion and the engine will begin to run rough.  ROP means Rich Of Peak.  Most carbureted engines have less even fuel distribution to each cylinder and the leanest cylinder will begin to mis-fire before an EGT drop is usefully noted. This is a real rough cut on the topic and there is a WHOLE lot more to this- look for articles by John Deakin (sp?) on AvWeb for better info than you found here!

  

 Mark



See what's new at AOL.com and

_________________
From The PossumWorks...

Float Flyr

Joined: 19 Jul 2006
Posts: 2704
Location: Campbellton, Newfoundland

Posted: Tue Oct 16, 2007 6:23 am Post subject: Knock sensors?

Thanks... I had heard of LOP before long enough ago to have completely

forgotten.



When in technical school we were always told to lean ROP 



Noel



[quote] --

_________________
Noel Loveys
Kitfox III-A
Aerocet 1100 Floats

frank.hinde(at)hp.com
Guest

Posted: Tue Oct 16, 2007 7:06 am Post subject: Knock sensors?

LOP is the most efficient place to run the motor. From my reading of the

Lycoming power chart it has about a 10% reduction in the BSFC in LOP vs

ROP (BSFC =.45 at 50F LOP and .5 at 80F ROP).



Rumour has it that the only way the old Howard Hughes Connies made it

across the Atlantic was to run the motors LOP. 



Its about the only place I run my motor apart from take off and landing.



Frank

RV7a TMX IO360



--

MONTY(at)bpaengines.com
Guest

Posted: Tue Oct 16, 2007 7:21 am Post subject: Knock sensors?

The Wright 3350 engines in Super Connies and also the B-29 were operated

LOP.  The FE had the most important job on the flight crew.  As I

remember,

The FE also had a set of throttles.  



Monty Barrett



--

Display posts from previous:

	Matronics Email Lists Forum Index -> Engines-List	All times are GMT - 8 Hours Goto page 1, 2 Next
Page 1 of 2

Jump to:

You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum
You cannot attach files in this forum
You can download files in this forum