Cummins head instrumented for in-cylinder combustion data

ENafziger

New member
Joined
Oct 22, 2008
Messages
843
Not sure where to ask this... but does anyone have info on where to procure an instrumented head for either a 5.9 or 6.7 Cummins? Preferably all 6 cylinders would accept Kistler or similar type transducers. This is for a research project where it's necessary to look at in cylinder combustion data.

Or...has anyone attempted this before that would share information on where to drill?

Thanks! --Eric
 
Is it for short term data? We use them on the bigger diesel engines (1000L+ displacement) and have had issue with reliability in Diesel and Heavy Fuel operation. The Natural Gas application has gotten much better but out of 12 engines (18cyl.) we change a couple a month.

if you can make it work the information you can get is amazing in terms of combustion heath and tuning. They certainly aren't cheap though.
 
Is it for short term data? We use them on the bigger diesel engines (1000L+ displacement) and have had issue with reliability in Diesel and Heavy Fuel operation. The Natural Gas application has gotten much better but out of 12 engines (18cyl.) we change a couple a month.

if you can make it work the information you can get is amazing in terms of combustion heath and tuning. They certainly aren't cheap though.

For a research project. It will only be run in a dyno cell...won't ever see on-road use. Looking for something to use for up to 24 months.
 
Do you already have the sensors and just looking for a place to add them? I really can't help with where to drill. More just curious what you had in mind for the sensors.

What I have been using are $1,500.00 sensor and that doesn't include the modules and software to read the data. I think it would be a bit over kill. there are some stand alone systems that used to be from a company called Diesel Dotcor or something like that. I know the sensors were Kistler also. I tired to find a link but didn't have much luck. I will be following the conversation.
 
Do you already have the sensors and just looking for a place to add them? I really can't help with where to drill. More just curious what you had in mind for the sensors.

What I have been using are $1,500.00 sensor and that doesn't include the modules and software to read the data. I think it would be a bit over kill. there are some stand alone systems that used to be from a company called Diesel Dotcor or something like that. I know the sensors were Kistler also. I tired to find a link but didn't have much luck. I will be following the conversation.

Possibly...depends on how things shake out. We could potentially share with another project in the cell if needed. They're the 6125B Kistlers IIRC. And yes, not terribly cheap. A set of 6 plus charge amps is over $20k. That cell has AVL DAQ on the software side.
 
Jon McAlravey from over at DD made a cyl pressure monitoring system for the Dmax. Its possible he could modify it for Cummins use. Hell, might be worth giving him a call and asking. 412-638-9026
 
Vaguely related, but isn't audi running glow plugs augmented with pressure sensors?

Something smart goes here
 
What I have been using are $1,500.00 sensor and that doesn't include the modules and software to read the data.

Back in college we had sensors built into spark plugs that were around $1500 each as well, and they definitely didn't last very long.

You're going to have to find a place to tap through the head too since I doubt you'll be able to get a sensor built into an injector.
Your readings will be different depending on where in the cylinder you have the sensor and how many readings per second it can pick up (plus it's a sensor and they're relatively slow to react).

What exactly are you trying to read? Pressure, temp, understand flame front, understand how the fuel's moving?

This is the type of testing that engine companies spend big bucks on typically.
 
Back in college we had sensors built into spark plugs that were around $1500 each as well, and they definitely didn't last very long.

You're going to have to find a place to tap through the head too since I doubt you'll be able to get a sensor built into an injector.
Your readings will be different depending on where in the cylinder you have the sensor and how many readings per second it can pick up (plus it's a sensor and they're relatively slow to react).

What exactly are you trying to read? Pressure, temp, understand flame front, understand how the fuel's moving?

This is the type of testing that engine companies spend big bucks on typically.

There is a very interesting graph I have seen on several occasions using this sensor, that compare cylinder pressure to crank angle. If you can get the sample rate fast enough you can see pre-combustion as it happens. No better way to tune timing in any given situation. Most of that data I have seen is test bed info. I have the ability to trend similar data at the plant now but I am not able to compare it to crank angle. You can even see the pressure spike and erratic combustion that causes the conventional injected engines to be more noisy than a common rail. I have often wished I had the ability to see this information in my truck.

We also use Bosch piezo knock sensors (looks like from a car) to measure the shock wave of pre-combustion. it works but not near as accurate at the pressure sensor. This may be an option since you simply bolt these on to the head directly. No need for a pressure port to be drilled in the head.
 
Not sure where to ask this... but does anyone have info on where to procure an instrumented head for either a 5.9 or 6.7 Cummins? Preferably all 6 cylinders would accept Kistler or similar type transducers. This is for a research project where it's necessary to look at in cylinder combustion data.

Or...has anyone attempted this before that would share information on where to drill?

Thanks! --Eric

I do this all day every day as I am a calibrator for an OEM. I will ask around and see what some people know. Most are out on vacation until the beginning of the year, I will see what I can find out then.

There is a very interesting graph I have seen on several occasions using this sensor, that compare cylinder pressure to crank angle. If you can get the sample rate fast enough you can see pre-combustion as it happens. No better way to tune timing in any given situation. Most of that data I have seen is test bed info. I have the ability to trend similar data at the plant now but I am not able to compare it to crank angle. You can even see the pressure spike and erratic combustion that causes the conventional injected engines to be more noisy than a common rail. I have often wished I had the ability to see this information in my truck.

We also use Bosch piezo knock sensors (looks like from a car) to measure the shock wave of pre-combustion. it works but not near as accurate at the pressure sensor. This may be an option since you simply bolt these on to the head directly. No need for a pressure port to be drilled in the head.

It's not that hard to convert to crank angle. 3000 eRPM is 18000 hz.
 
Real-time cylinder pressure must be the holy grail of engine tuning.

I imagine you essentially figure out that your head gasket is good for X PSI, then adjust timing, fuel and air to get as close to X PSI at as many points as possible?

A head with glow-plug ports would seem like it'd make transducer installation easy. I don't think they ever made 6BT's with glow plugs though...
 
Last edited:
Eh, the head gasket really isn't considered, at least not on production engines. On a diesel, it's really for finding mean best torque timing. And, for making sure that the cylinders are all withing x.xx% of each other. I use it at work to monitor to health of the engine as I am completing the calibration process. Mean best torque timing is typically determined pretty early in the development process and then left alone, unless a hardware change happens (cam, intake, exhaust manifold). Typically, transducer placement is as close to the middle of the cylinder as possible.

On a gas engine, pressure transducer spark plugs are/ can be used. I have also seen where a very small hole is drilled diagonally to the middle of the combustion chamber and then the pressure transducer is threaded into the cylinder head.
 
Good luck. You would be better off making friends with someone at Cummins who could get ahold of a head that's been drilled for a transducer.

Going to need a crank encoder to get precise readings. Mark at Danville has had great results using cylinder pressure monitoring to dial in big injectors and pushing the limits of stock engines.
 
I might have pictures of a transducer in a 6.7 head.

Sent from my XT1093 using Tapatalk
 
I do this all day every day as I am a calibrator for an OEM. I will ask around and see what some people know. Most are out on vacation until the beginning of the year, I will see what I can find out then.

Thanks...that'd be appreciated.

To be more specific for those interested, the sensors we typically use are these (or similar). They are specifically made for in-cylinder combustion work: https://www.kistler.com/?type=669&fid=54410&model=document&callee=frontend

What exactly are you trying to read? Pressure, temp, understand flame front, understand how the fuel's moving?

This is the type of testing that engine companies spend big bucks on typically.

I understand it's not cheap, but we have the supporting infrastructure in place...my question was mostly on where to source a head that was drilled and tapped for the sensors. Or if anyone had information on where the head could be drilled.

Obviously, these heads don't have glow plugs, so it necessitates drilling a new hole. I've never tried to do this on a 5.9/6.7 head, so looking for information before we start experimenting. Sometimes it can be tricky to miss water jackets, etc. Casting shift and variations can also make the job more challenging. I figured as long as these engines have been around, that there may be a chance someone has "been there and done that" and could provide some info.

We also use Bosch piezo knock sensors (looks like from a car) to measure the shock wave of pre-combustion. it works but not near as accurate at the pressure sensor. This may be an option since you simply bolt these on to the head directly. No need for a pressure port to be drilled in the head.

Interesting...I'll look into specs on these.

Real-time cylinder pressure must be the holy grail of engine tuning.

Yes, it is! Typically this type of work is done in conjunction with a quadrature encoder on the crankshaft; we typically use 1800 pulse/rev to give 0.2° resolution.

This is the software that we're moving to on the newer setups: NI Combustion Analysis System Software for NI LabVIEW - National Instruments

In simple terms, it let's you look at things like:
  • Pressure trace...you can see the in cylinder pressure real time with respect to the crankshaft. Obviously max pressure is of interest, but the shape of the curve gives alot of info as well
  • PRR...pressure rise rate...the "rate" at which the pressure initially rises due to combustion is paramount. Usually 10 bar/degree is considered the upper limit for production; in the research environment you may see 10x that (but not for long)
  • IMEP...indicated mean effective pressure...the "indicated" work that is being done on the piston
  • MFB...mass fraction burned...common points are 5, 10, 50, and 90 percent points
  • HRR...heat release rate
  • Cylinder balancing...presuming you have multiple cylinders instrumented. When looking at advanced combustion modes, balancing becomes critical as you push things to the edge of stability

Yes, but the real trick is know exactly where TDC is...

This can be done by physically aligning the pulse on the encoder with TDC using an oscilloscope. But we have a capacitive probe from AVL that allows you to do this dynamically with greater accuracy. Max pressure in the cylinder doesn't occur exactly at TDC.

I might have pictures of a transducer in a 6.7 head.

This could be helpful!

Thanks for the help everyone...

--Eric
 
ORNL still?

I can put you in touch with a guy that knows some guys at cummins if that would help.

Is this all static load testing?
I've always wondered how power train load variation affects combustion characteristics.

Something smart goes here
 
I've seen some engines where they tap the block, so the sensor is right above that first compression ring. Yes, it's a disposable block at that point, so not sure that is the right answer for your setup.

I would love to see the data collected.
 
Back
Top