2016 Proposed ST Rules Revisions

[Greg Smith]

As far as a varience, I agree the amount of cushion should be left up to an individual competitor's risk tolerance.

No need to further complicate things with another calculation.

Have you ever been dyno'ed at the track? If so, how did the numbers compare to your off-site dyno? "Another calculation", really? Because multiplying something by 1.02 is just too complicated .

[J_Roberts]

When I was recently checked for ST2 class compliance, the Group Lead was using the NASA ST#/TT# class calculator available on the NASA web site; that is not a standalone app, but just a MS Excel spreadsheet. I had my Class Calculator form filled out of course, which I had calculated by hand. So all he had to do was enter my inputs, and check that we had the same results.

 

I know this was suggested earlier, but why not make a similar, official, downloadable spreadsheet for calculating the average HP as well? Then, you can use more than the three data points suggested, and/or something more sophisticated, and fit the competitor's power curve more precisely, making it harder to game the system. All the user has to do is input the RPM vs HP table values into the spreadsheet, and let it calculate average HP per the formula (whatever that may be). I think it is reasonable to assume the Group Lead will have a laptop at the track, so they could have the 'official' spreadsheet ready to go, no need for cell phone reception. Good point by the way, some of the tracks are in areas that have crap cell phone reception, like Eagles Canyon in Texas.

 

With a spreadsheet, now you are free to use a more complex formula that will better level the playing field for high-RPM vs NA vs turbo vs whatever engines, without requiring a mathematician for the calculation. For example, this could allow the use of an approximated area under the curve method, and/or compensation for the RPM range of that engine as pointed out (very well BTW) by BJ Meyer earlier in the thread. Either way, all you are doing from a competitor/compliance check standpoint is typing in the HP vs RPM values, which is easy. Make the formulas available, and the calc steps, so if someone wants, they can do it via a hand calc. But, the Group Lead at the track still uses the official spreadsheet tool for the compliance check, so it is easy for them. I'll bet if you polled all of the NASA ST/TT Group Leads nationally, most (or all) already take laptops to the track anyway.

 

It could even be integrated into the current ST#/TT# class calculator, so that it is all done at once.

Edited November 10, 2015 by Guest

[stevensa]

I like the new avg HP calculation methods. Simple and fair.

[Sterling Doc]

As far as a varience, I agree the amount of cushion should be left up to an individual competitor's risk tolerance.

No need to further complicate things with another calculation.

Have you ever been dyno'ed at the track? If so, how did the numbers compare to your off-site dyno? "Another calculation", really? Because multiplying something by 1.02 is just too complicated .

 

It's not hard, it's useless. It just sets the cap 2% higher, and everyone tunes to that. If you are comfortable with 2%, tune to 98%. Others may want 3%. Decide for yourself, and do the math once.

 

I wrote the 944 spec Dyno rules. People wanted an dyno variance so we wrote the rule as 138+2 (for Dyno variance). HP+TQ/2. 3 years later, all people remember is the cap is "140". If anyone remembers it's when they go over and claim they tuned to 138. Doesn't matter, you're still over. Useless.

[kbrew8991]

You may want 2%, I may want 5% - pick your own variance

[Greg G.]

 

EDIT: For the love of god the abbreviation for average is avg, not ave! Sorry, that's been bringing my OCD out.

Don't be such an extremist

While the government does recognize Avg. as the scientific abbreviation for average, AVE is common:

http://www.abbreviations.com/term/124050

http://www.allacronyms.com/AVE/Average

http://www.acronymfinder.com/AVE.html

http://acronyms.thefreedictionary.com/AVE

[Greg G.]

Wow deja vu!

David, this was one of our original formulas that we had worked out pro formas for. We didn't use it the first time, because of the 10 versus 3 extra data points that were needed. We thought we could get away from the increased number but obviously couldn't with the ability that the tuners have. Thanks for bringing this back up, though.

[Greg G.]

I'm okay with this proposal, thanks for David Farmer for the idea a couple pages back. I do think there needs to be specific wording on how precise we need to be with the +/- 500rpm intervals. For instance I make peak HP at ~8350RPM, so do my -500 rpm intervals need to be at 7850, 7800, or 7900RPM? For the peak number are we going to use the peak number on the graph, or the peak number on the data export? For instance my peak shows 376.56 HP on the graph, but 375.33 HP on the data export with 100 RPM intervals.

Greg, we will use the Peak HP number on the graph. The intervals will be at least to the nearest 50 rpm. I'll get back to you guys on this, as we are checking to see exactly what we can get out of the standard winpep program that all Dynojets have, and if we can get anything more customized for NASA.

[Greg G.]

I would also like to propose a dyno variance of 2%, similar to that in GTS. The way I understand the current rules is if you're anything over your declared dyno number you would be DQ'ed, dyno's just aren't that accurate and there needs to be some allowed variance.

The other guys are correct. Adding an amount for Dyno variance is just moving the bar. As well, it actually sets drivers up for possible failure if there is no allowance to declare a higher HP number than was obtained during pre-competition testing. If the driver knows that his vehicle tends to show a lot of variance on the Dyno (not because of the Dyno, because of the variables inherent to his vehicle), and he is limited to just whatever variance amount the rules state is permitted, he is always a dice throw from a DQ. As Eric stated, if one has a Dyno limit by the formula of 300 hp, and then someone tells him that he has a 2% variance allowance, for 90% of the drivers, their brain immediately goes to, "Oh, my Dyno limit is 306 hp." Then, some will say, I'll tune for 303 hp because my car is always spot-on on the Dyno....etc. The best way is to let the competitor who knows the vehicle, and who is the one responsible for making sure that the same amount of fluids are in the vehicle, at the same temperature, with the same tires/wheels/pressures, with the same cleanliness of the air filter, etc. be the one to decide on how much variance he needs to build in to prevent a DQ.

[sperkins]

Guys, try this one--latest draft proposal in the comment phase by Directors (and now you!)

 

New for 2016: The number used for calculating the Adjusted Wt/HP Ratio for the ST3 (and TT3) classes will no longer be the maximum horsepower of the three Dyno runs. It will be a calculated average giving a better approximation of the maximum horsepower available over a range of usable RPM. The Dyno test with the highest maximum horsepower will be used to calculate this average (not an average of the three Dyno runs) as follows:

 

Ave HP = Average HP calculated and used in in the Adjusted Wt/HP Ratio

Max HP = Maximum horsepower

 

The following ten (10) data points will be obtained from the Dyno’s numeric RPM/HP table printout:

Horsepower at: 500 rpm, 1000 rpm, 1500 rpm, 2000 rpm, 2500 rpm greater than Max HP rpm

Horsepower at: 500 rpm, 1000 rpm, 1500 rpm, 2000 rpm, 2500 rpm less than Max HP rpm

(If any of the above data points at higher RPM than Max HP RPM do not exist due to redline, then those potential data points will not be used in the calculation of Ave HP.)

 

The highest three (3) data points of the above ten (10) will be used in the calculation below:

 

Ave HP = Max HP +(sum of the highest three data points) Divided by 4

 

 

Looks good. Hopefully this will carry over to at least ST2.

[TWI]

Has any consideration been given to keeping the original TT3/ST3 Wt/HP ratio since this up to 10pt average calculation will produce a higher number than the initial proposed 3pt calculation?

 

Edit - after playing with a little bit of the math, I predict that there will be some real interesting dyno curves tuned into the top TT3/ST3 cars.

[davidfarmer]

that was the point...the 3 pt version was very easy to game for some cars. The best of 10 will be much harder. I suggest requiring dynojet data at 100rpm increments going forward. It would be very hard to hide any useful data in 100rpm chunks!

 

 

 

[bpanther]

The following ten (10) data points will be obtained from the Dyno’s numeric RPM/HP table printout:

Horsepower at: 500 rpm, 1000 rpm, 1500 rpm, 2000 rpm, 2500 rpm greater than Max HP rpm

Horsepower at: 500 rpm, 1000 rpm, 1500 rpm, 2000 rpm, 2500 rpm less than Max HP rpm

(If any of the above data points at higher RPM than Max HP RPM do not exist due to redline, then those potential data points will not be used in the calculation of Ave HP.)

 

The highest three (3) data points of the above ten (10) will be used in the calculation below:

 

Ave HP = Max HP +(sum of the highest three data points) Divided by 4

 

I like it, let's give it a shot in 2016. It's a bit more of a pain to do the math, but not that hard. In RM, I think we will require the competitor to supply their .dlf file. I don't want to guess with a ruler.

[Eric W.]

 

I like it, let's give it a shot in 2016. It's a bit more of a pain to do the math, but not that hard. In RM, I think we will require the competitor to supply their .dlf file. I don't want to guess with a ruler.

 

Dave, you can also have them print the data. File->ExportData>Print.

 

You can select 500RPM increments and it'll spit it all out in a table.

[Greg Smith]

 

I like it, let's give it a shot in 2016. It's a bit more of a pain to do the math, but not that hard. In RM, I think we will require the competitor to supply their .dlf file. I don't want to guess with a ruler.

 

Dave, you can also have them print the data. File->ExportData>Print.

 

You can select 500RPM increments and it'll spit it all out in a table.

Keep in mind it's 500 increments based off your highest HP RPM which more than likely doesn't fall at x000 or x500 RPM.

[loftygoals]

I suggest requiring dynojet data at 100rpm increments going forward.

 

I've been playing with the collection of dyno files I have from various cars in WinPEP 7, DynoJet's software for reading and working with the .drf run files. The nice thing about WinPEP is that it is free to download, so if you get the run files from your dyno shop you can export the data yourself.

 

Anyway, the thing I'm noticing is that I often don't see the peak horsepower from the graph listed in the report. At 100 RPM increments, peak from the report can be off be low by over 1 hp and in some cases shifted by a somewhat significant amount of RPM. The distance from the desired RPM data points and a points listed on the report can be distant as well. This distinction is important, because we need our data points relative to peak horsepower. As the proposed rule is currently written, if peak is @ 6273, the increments should be (..., 5273, 5773, peak, 6773, 7273, ...). Since the reports are generated in even increments, we will have to use small increments to increase proximity to the desired RPM.

 

I think Greg's suggestion of 50 RPM increments gives acceptable resolution. At 50 RPM steps, I'm seeing within .5 hp of peak hp and I end up much closer to the desired data points on the report. At smaller than 50 RPM steps, you end up with too much data to dig through.

 

 

Also I was currious about how it will be determined which dyno run we use? Currently, we use the highest peak of three:

The highest peak horsepower number of the three tests will be used as the official certified horsepower for weight to horsepower calculations.

 

Will this continue to be the case, or the highest average?

 

-bj

[brkntrxn]

To not over complicate it, can we just use "the maximum peak horsepower from the 100rpm increment data" instead of "maximum peak horsepower"?

 

If I have a peak that occurs in between 6100 and 6200, how much does that extra 1,2,5, 50 hp really do for me if it occurs at one small blip on my rpm scale? It may be possible to do and have a dyno chart that looks like a sawtooth sinewave, but it isn't usable. If my competition wants to spend gobs of money for a gnats @$$ worth of extra power that is going to drive his cylinder flow and exhaust temps nuts, so be it. I'll be racing while they are tuning.

 

If we keep the data measures to the 100rpm increment lists that you guys are saying can be exported from a Dynojet, that will make it much simpler. I could really care less if my competitor actually has 5 more peak horsepower in between the solid 100rpm increment marks from his data, because I'd rather have easy compliance for everyone.

[bpanther]

 

I like it, let's give it a shot in 2016. It's a bit more of a pain to do the math, but not that hard. In RM, I think we will require the competitor to supply their .dlf file. I don't want to guess with a ruler.

 

Dave, you can also have them print the data. File->ExportData>Print.

 

You can select 500RPM increments and it'll spit it all out in a table.

 

I'll ask for the full file so that we can do it and as needed. And fix the smoothing. And turn on correction. And do it at 100 HP intervals.

[loftygoals]

To not over complicate it, can we just use "the maximum peak horsepower from the 100rpm increment data" instead of "maximum peak horsepower"?

 

It's not the peaks, but the data itself. I just looked at a BMW M3 dyno I have. It makes approximately 250hp at peak (natural curve, not detuned). It's only 3.2L (S52). Sampling the data in 100 RPM steps, I see as much as 9 hp change between steps. So if I shift my data points by 100 RPM's I could see my average power increase or decrease by as much as 9 hp. 9 seems to be an outlier, though. The typical change seems to be 4-5hp on motors of similar size and power. So if the variance could be +-5 horsepower, you could have two competitor both at the legal max that have an actual 10hp difference in the average. It should also be noted that an increase of 10 Avg HP is much more significant than when you are talking about just 10 HP at peak.

 

My point is the resolution does matter. If you take it to 50 RPMs, the typical change will be halved. This also halves the variance to a max of +-2.5hp per car or 5 hp between cars, which I think is very reasonable.

 

-bj

[zdr93523]

I really like the new average HP calculation. I think this will help equalize competitors with various power curves.

[J_Roberts]

After reading some of the previous posts, and playing around with some power plots in Excel, I agree. There should be more resolution (than 500 RPM increments and three non-peak data points) used for the average.

 

With that level of fidelity, there can be a significant difference in area under the power curve for two power plots, where both have the same average power calculation per the proposed rule; I am looking at this comparing a "normal" V8 non-electronic-throttle kind of plot, and a FI digital-wastegate-controller (or throttle by wire) kind of plot, comparing the two in Excel.

 

As pointed out, the finer the RPM resolution, and the more data points used in the average, the harder it will be to electronically game the system, and the more representative the correction will be.

[davidfarmer]

Greg not trying to steal your thunder....it's a great system and not too complex

 

50rpm print out from your Dynojet, peak plus next 3 highest points in 500rpm jumps above and below the peak for 5000rpm. This doesn't hurt those of us with flat curves, and it should really help those with peaky engines.

[Greg Smith]

At 100 RPM increments, peak from the report can be off be low by over 1 hp and in some cases shifted by a somewhat significant amount of RPM.

I think Greg's suggestion of 50 RPM increments gives acceptable resolution. At 50 RPM steps, I'm seeing within .5 hp of peak hp and I end up much closer to the desired data points on the report. At smaller than 50 RPM steps, you end up with too much data to dig through.

I'm for using 50 RPM data export increments and use the highest peak HP number from that export(versus the one from the dyno graph which can be ~1hp higher) for ease of calculation. There should also be a contingency if the same peak HP occurs at 2 different points on the dyno that the equation is calculated for both peak HP points and the one with the higher average will be used.

[loftygoals]

There should also be a contingency if the same peak HP occurs at 2 different points on the dyno that the equation is calculated for both peak HP points and the one with the higher average will be used.

 

Solid catch of a gotcha and good a resolution to boot!

[farberio]

50rpm vs 500rpm will have a large effect on flat torque engines because the closer the resolution the smaller the extra hp a peaky engine will get.