Anyone ever tune with in cyl press. measurement ?

[Yves]

I'm seriously considering buying an in cylinder pressure sensor at the moment. However I would need to build my own programs to analyse the data.

Apart from the time it will take, the cost is considerable for just one sensor and adapted spark plug. So I wonder how good the results would actually be and whether it is worth the investment.

Thanks

[billr]

My opinion? The money and effort would be better-spent on dyno time. Some factors, like rod-length/stroke ratio and piston-pin offset are easy to define in your analysis program; but friction due to the side-loading on piston/cylinder is probably still going to be quite a guess. Measuring output torque kind of "integrates" all that stuff automatically.

[Yves]

The objective here is not to measure power.

[billr]

What is the objective, then? Are you only trying to measure peak combustion pressure? That is only one factor in tuning an engine.

[Yves]

optimum ignition timing/detonation.

Have you ever tuned with one ?

[billr]

No, I have not.

What will you be optimizing ignition timing for, if not torque/power? Setting advance to maximum possible before detonation is no assurance that best power, fuel economy, or emissions has been achieved.

[motthomas]

What objectives are you trying to achieve with in cylinder pressure measurement? What are you trying to measure with it?

In my experience it is mostly used to collect data and verify certain parameters and not as a primary method of calibration.
For instance, MBT timing is dialled in on a dyno and then verified with the pressure trace. Knock is detected & controlled using acoustic knock sensors while cylinder pressure traces are used to quantify amplitude & nature. It is useful to determine individual cylinder filling distribution but EGT will also tell you that.

Maximum cylinder pressure is a useful measurement alright but unless you know the maximum pressure that can be run without damaging your specific pistons, rods & bearings it is simply another piece of information.

As with most tools, unless the cylinder pressure trace measurement is accurate, it is no real use as a tool. The industry standard kit uses a dedicated 720 tooth optical crank angle encoder giving the software a physical reference ever 0.5deg. Without that kind of resolution the pressure trace is just not accurate enough to be useful.

Don't get me wrong, it would be an awesome project to try and do it but you need to be very clear in your goals and why you think you need this equipment. It is certainly no going to be the holy grail, fix all your calibration issues tool.

Sent from my GT-I9305 using Tapatalk

[Yves]

We all know that optimal peak pressure needs to occur in the 10 to 15 degree window atdc. So if you can get say 10 degrees without any knock trace I think you will be very close to mbt.

[motthomas]

I hadn't seen the earlier replies before posting the above so I will just add a little to it.

If your goal is simply to acheive optimum ignition timing then this can be acheived far cheaper and more reliably on a dyno. Optimum ignition timing is the later of either MBT or borderline knock, it is no more complicated than that. Even a dyno that is way out of calibration will allow you to find MBT timing as all you are looking at is a relative torque measurement. To do the same using in cylinder pressure measurement you need to be 100% confident that your amplifiers, computing hardware and software are calculating the correct thermodynamic formulae to provide you with an answer that is reliable. There is a reason why the industry uses one companies hardware and software and that company can charge close to the cost of a house for the required hardware and software. That doesn't even include the pressure transducers.

The idea is admirable but I think you are underestimating the complexity of getting such a system to provide reliable data and overestimating the advantage it will give you over other more common methods of calibration.

[Yves]

Have you tuned with it ?

A dyno here costs roughly € 1000 a day. Change something on your setup and you're going over there again.
Cost of the sensor and daq is € 2000 and the rest is time invested.

[rukavina]

How about finding a LONG straight road, step on throttle, go fast,check time and do again....... lot more fun

[jsmcortina]

Even with the daq you still really want a dyno to safely and repeatably use it and validate the results.

James

[motthomas]

Have you tuned with it ?

A dyno here costs roughly € 1000 a day. Change something on your setup and you're going over there again.
Cost of the sensor and daq is € 2000 and the rest is time invested.

Yes. I use it every day. However, as mentioned in my previous posts it is used more as a data aquisition tool for combustion analysis as opposed to calibrating. Sure, peak cylinder pressure occurs around 15°ATDC at MBT timing and there are other correlations can be made using combustion phasing and rates of pressure rise but none of these are exact. The rules of thumb are derived from empirical data as opposed to actual theory and they do differ slightly from engine to engine. The only way to positively identify MBT timing is using torque measurement on a dyno. The one advantage cylinder pressure traces will give you is an indication of how far away from MBT timing you are running on a knock limited engine but then what are you going to do with that information? On a knock limited engine you can only run to the knock limit anyway.

The other thing about calibrating from combustion data is that it is by nature quite variable so in order to make any sense out of it, you need to be running the engine with all other parameters tightly controlled under steady state conditions for a period of time (at least 300 cycles) so that the combustion data can be averaged over that time. You will not be able to acheive this level of control on the road and so, as James said, you will need to get on a dyno anyway. Also, if your intention is to only use one transducer in a multicylinder engine, that will not provide you with a very good representation of what the entire engine is doing and unless you can very precisely replicate the engine running conditions over and over again, you will not learn anything by swapping the one transducer around between cylinders.

I understand that dyno time is expensive in Belgium and that must make things more difficult for you. If you plan what needs to be done on the dyno though you can minimise the time needed to dial in a setup. Realistically you should only need to spend a full day mapping the engine once. Once you have a good base calibration, even if you make a change (unless it is a drastic change) one or 2 hours on the dyno should be enough to adjust the calibration to suit. Using a planned approach should allow you to base cal an engine and allow for c.4 changes to be made before you even get up to the cost of a transducer and daq and it will ultimately be a better calibration.

[Yves]

Ok, that's usefull information. Thanks.