PRI Show

[fishracer]

I am new to this site and find it very informative.

I also recently attended the PRI Trade Show in Orlando and saw some interesting products while there. One company had a boring head that was programmable & expanded as it machined below the minor diameter. It appeared to be a high quality piece and is targeted to cylinder head technicians.

Another company had an inertia type engine dyno with a series of large diameter, heavy flywheels. It also had a Stuska absorber connected to the main centerline. I do not fully understand the concept and was hoping that someone here could explain it to me.

Thanks.

 

[fishracer]

Sorry, But I'm not understanding your point. Are you thinking that our interest in this dyno is for complete track simulation?

Our interest is only for engine-acceleration testing. Not for durability testing and/or endurance testing. We have no need to exactly simulate deceleration, gear changes, etc. (although we will be doing some part-throttle and off-throttle testing in addition to WOT testing). Sorry for the confusion.

 

[patprimmer]

Someone said these dynos can exactly replicate real world. I am making the point that despite considerable improvement it still is not exact. Same goes foe effects of inertia on fuel system and oil system

Regards

eng-tips, by professional engineers for professional engineers
Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

 

[NCShane]

It sounds like this dyno would be pretty good for back to back component or set-up comparisons. Will you be getting the on that has a water brake attachted to it also? That would give the ability to do your initail calibration at steady state.
As for the filtering question, our ac dyno has four outputs for speed and torque
1. Raw data
2. a 1hz filtered and averaged signal used for feedback control
3. a 1hz 4th order Chebychev filter
3. a 10hz filter for displaying stable readings to the user

I log the raw data and filter it in Excel, etc depending on the desired final product. The level of filtering depends on your sample rate. Our software can record at 4167 Hz, so it takes a good bit of filtering to make a nice power curve. If I am doing a power curve on an engine that is already calibrated I reduce the sample rate to about 20 Hz to clean up the data.

 

[fishracer]

You're correct - no dyno can exactly replicate a race car on a track.

I took the message of the engineer at this dyno company to be that the inertia dyno could replicate the acceleration characteristics of an engine on the race track more closely than a brake style dyno.

I did a poor job in conveying his message!

 

[bsfc9]

This is a very interesting thread. I see two different subjects being discussed here. I am interesed in the data aquisition filtering subject. In particular, is anyone familiar with the DEPAC dyno aquisition program. It claims no data sampling and subsequent filtering. It uses "Smart Averaging" which is some sort of mathmatical integration routine to show data. It claims the most accurate and "The Best Man Can Do" according to staticical theory. Thanks in advance.

 

[NCShane]

I don't know anything about the DEPAC stuff, but I took a look at the website this morning. It doesn't look like they are doing anything unique with their controls system. They don't seem to give any details about their "Special Statistical Process" but I'm not sure how any processed data could be more accurate than raw RPM and torque data. They have some interesting info on the website though.

 

[SomptingGuy]

fishracer: My comment was slightly [OT]. The kind of "raw data" I like to have is the stuff that comes from high frequency sensors (pressure transducers, microphones, accelerometers, inductive probes, etc). I don't have "my own program" as such, just Matlab and techniques I've learned over the years.

 

[bsfc9]

Fishracer: sorry to hijack you thread.
NCShane(or anyone for that matter): Check out this info the the DEPAC "special statistical process". Let me know what you think--thanks.

http://patft.uspto.gov/netacgi/nph-...4,870,585.PN.&OS=PN/4,870,585&RS=PN/4,870,585

 

[fishracer]

SomptingGuy: I understand. Have you successfully used Matlab with any raw dyno data?

bsfc9: I am pretty familiar with the DEPAC system and it is very good. However the "smart-averaging", like filtering is utilized to smooth the data enough to practically analyze it. (I use the term 'filtering' as a blanket description for any manipulation of the raw and/or real data, which would include any "averaging".) Each dyno software company utilizes a filtering process. They use different methods, but all do it. There is too much "noise" involved in testing a race engine and the filtering is necessary to smooth the data. ("Noise" from rf, power pulses, ignition scatter, etc.)

The problem comes in when a dyno software utilizes excessive amounts of filtering in order to manipulate the data to an extreme. (I've witnessed dyno [acceleration] tests that sound like a stepped test - yet when the graphed data shows on the screen, it is very smooth, rounded, etc. and looks nothing like the run it came from. In my estimation - this is too much filtering! That particular dyno-company does not allow the customer access to the raw data!

This is one feature that is attractive about the inertia dyno. Since the engine is accelerating a fixed mass (just like a car) the acceleration run is very smooth and the data does not need the excess filtering to smooth it.

 

[fishracer]

NCShane:

Yes, we will get the dyno with the water brake attached.

.75 Hz is the amount of filtering they were using on the rpm channel of the inertia dyno, no filtering on the elapsed time. I did not check any other channels, but all are user defined as to the filtering %.

The software uses a 3000 Hz board and up to 1000 Hz per channel.

What brand of AC dyno do you own?

 

[NCShane]

bsfc9: the DEPAC patent info is interesting, I have a few comments, but I want to finish reading it first

fishracer: We have a Schenck Pegasus AC unit.

 

[fishracer]

NCShane: Is the Schenck unit a chassis or engine dyno?

 

[NCShane]

It's an engine dyno. It's a 460 Hp 8000 RPM unit with GE drives and Schenck controls. Unfortunately the software is a bit outdated, but Horiba bought out Schenck and no longer supports the dyno control equipment.