Dyno Dynamics - Shoot out modes

[Mignon]

Fascinating. I would like to know how close the algorithms come to my own equations or vice versa.

[PeterBurgess]

May pose a risk being generic rather than specific? I wonder if one can make tyre choice with the 'classes' mentioned?

[stevieturbo]

Mignon said:

Do I deduce from this that shootout mode goes straight from wheel bhp to flywheel bhp via transmission and tyre loss equations rather than using any sort of coastdown procedure?

Pretty sure DD shootout mode offers only wheel numbers. The operator would then need to request guesstimated flywheel figures after that

So trying to use a flywheel figure would not be normal use of shootout. US, OZ etc rarely try and offer flywheel figures, it's an odd UK thing, presumably so people get higher numbers for the pub.

[PeterBurgess]

LOL Stevie, we used to call them pub bhp On our old Clayton with no print out the bhp went up as they went down our entrance ramp for some folk. In fact Wayne that used to work for us was in a pub up in Alfreton and heard a customer talking about the dyno we had done and the bhp was well in excess of what we had measured!
On our new rollers we have a cush drive between the pau and the roller, there is a small movement in this which means, if there is a hiccough in the power delivery when we start a power run, the pau momentarily loses contact with the roller so we get a 'pub talk' torque figure of double or even triple
I still like the coastdown test so we can directly compare losses with same type so we can see if there are problems, often given away by billows of smoke coming out of a wheel arch!!!! I can see the appeal of the mathematically derived figures with the Dynodynamics but it would be nice if they were coupled with a coastdown test which would have an effect on the 'realism' of the final figure.
Karl and Oli are wary of shoot out modes especially with 4wd Cossies which sap power like there is no tomorrow, I wouldn't be surprised to see 100 bhp gone to drive the transmission and tyres.

[spitfire4v8]

You can do a coastdown on the DD but I'm not sure I know of anyone using it.

The DD modelling to get back flywheel seems very believable for the cars I regularly run and the numbers for nicely modified engines seem to correlate well with what Mignon suggests as the typical maximums for torque/litre as well ..

The argument over wheels figures becomes a moot point though when you consider dynos such as the dynojet give wheels figures around 15 percent higher than the dyno dynamics. Even the guy who invented the dynojet says it over-reads massively.
So .. wheels figures are great, but not all wheels figures are the same

In reality it's a minefield and there's no wonder the internet is packed full of rolling road arguments (I've been involved in many myself over the years), even the manufacturers can't agree on what's being recorded, what hope has the man on the street.
Bottom line I guess is the good advice of using the same dyno to plot the progress of any upgrades, and make sure the operator is consistent in his/her technique. And treat absolute numbers a guide only, comparison between dynos is going to get you in trouble.

there's one hub dyno operator I know who quotes his hub figures as wheels figures, and adds 17/18 percent on top. The industry is often more guilty than the man on the street at exagerating claims, or just not realising what's actually going on with his own testing.

[Mignon]

spitfire4v8 said:

You can do a coastdown on the DD but I'm not sure I know of anyone using it.

The DD modelling to get back flywheel seems very believable for the cars I regularly run and the numbers for nicely modified engines seem to correlate well with what Mignon suggests as the typical maximums for torque/litre as well ..

Which is here for those who have still not seen it.

https://web.archive.org/web/20110918114921/http://...

[Mignon]

spitfire4v8 said:

In reality it's a minefield and there's no wonder the internet is packed full of rolling road arguments (I've been involved in many myself over the years), even the manufacturers can't agree on what's being recorded, what hope has the man on the street.

It's certainly not surprising that the average punter doesn't really know what is being measured on a rolling road which was the inspiration for me to write my articles many years ago to try and educate them. It's more disappointing when the people using dynos also don't know what is going on or at least don't use any sort of critical thinking before touting their numbers. I once chatted to a very long time (30 years or more) rolling road guy who had no idea torque and bhp were mathematically linked or what either of them really was. In 30 years it had never occured to him why the graphs crossed at 5252 rpm, how you calculated torque from bhp and rpm or vice versa. He had a reasonable idea what power was and what sort of levels to expect from different engines but had no clue why the 110 ft lbs per litre from the normally aspirated car he'd just finished testing couldn't be correct. After some more investigation it turned out he'd "calibrated" his rollers by assuming the output from his old shagged out road car would be the same as the manufacturers flywheel ones and twiddling with the calibration screw on the dyno control until they were.

I know certain people in the tuning industry who are convinced the high 90s ft lbs per litre they regularly see from their own home devised correction from wheel figures back to flywheel ones is because they are just much better at tuning engines than anyone else. Better it seems even than the guys in F1 and Nascar who don't get as much torque per litre from custom designed race engines as they do from modified road ones.

[stevesingo]

spitfire4v8 said:

You can do a coastdown on the DD but I'm not sure I know of anyone using it.

The DD modelling to get back flywheel seems very believable for the cars I regularly run and the numbers for nicely modified engines seem to correlate well with what Mignon suggests as the typical maximums for torque/litre as well ..

The argument over wheels figures becomes a moot point though when you consider dynos such as the dynojet give wheels figures around 15 percent higher than the dyno dynamics. Even the guy who invented the dynojet says it over-reads massively.
So .. wheels figures are great, but not all wheels figures are the same

In reality it's a minefield and there's no wonder the internet is packed full of rolling road arguments (I've been involved in many myself over the years), even the manufacturers can't agree on what's being recorded, what hope has the man on the street.
Bottom line I guess is the good advice of using the same dyno to plot the progress of any upgrades, and make sure the operator is consistent in his/her technique. And treat absolute numbers a guide only, comparison between dynos is going to get you in trouble.

there's one hub dyno operator I know who quotes his hub figures as wheels figures, and adds 17/18 percent on top. The industry is often more guilty than the man on the street at exagerating claims, or just not realising what's actually going on with his own testing.

I wouldn't say over reads, they measure differently. Any dyno with a single large roller will have a better tyre/roller interface than an twin roller dyno due to only on contact patch and that when toque is applied through the tyre, it is not trying to climb up the front roller and lose contact with the rear roller.

I friend of mine and I built similar engines, 2.5lt 4cyl 16v. There were two differences in the engine, mine had 0.5:1 more compression and a ported head.

On a DynoJet, measured wheel HP was almost exactly the same 235 v 237 or something and he had much better low down torque. I was initially gutted and perplexed. Then I realised when comparing the graphs that the road speed of the other car was much lower. That car was tested in 3rd and had a shorter diff, much shorter; 3.64 vs my 3.15. My car was hitting 137mph in 4th (which isn't direct), he was hitting 87mph!

When I looked at the results of the coast down test, the losses on both cars were the same at 87mph, only thing was I had another 40mph to go to hit the red line where my losses were double the other car.

When comparing the flywheel power my engine showed significant power advantage above 4k rpm and matched the other engine below that.

So, IMO there is value in coast down tests if only as a comparator of relative losses, especially in twin roller dynos where over enthusiastic strapping down can lose chunks of wheel power which the algorithm can't account for.

This happened on the initial mapping of my car when it made much less than I expected. The dyno operator (whom was not invested in my car nor my type of car) was like, "oh well, not so good then eh!". When I questioned the smell of burning tyre rubber, he was like, "Yeah, sometimes get that. Not on the cars I tune though!"

The DynoDynamics algorithm can't allow for that, so I'm not a fan TBH.

[AW111]

We found that coastdown tests don't correlate well with actual losses, at least on big horsepower cars. The more torque the tyre is transmitting, the more it deforms.
Also, the tyre deforms as it tries to climb the roller against the straps, under power.

A large roller dyno has different issues with traction, especially once you get over 6-800 hp.

Over 1,000 you're going to have traction problems no matter what - buy a hub dyno if you're regularly tuning cars in the 1000 hp range .
As we get more hub dynos out there, we should get some more insight into tyre vs transmission losses.

Just to add : as a tuning tool, consistency is the most important thing. That includes how you tie the car down, tyre pressures, fan, ramp rate, etc.
The best dyno in the world is useless if you don't use it properly.

Conversely, I know a tuner who specialises in classics, with many repeat customers going back years. He knows his dyno is out of calibration, probably ever since he purchased it second hand. He doesn't want it recalibrated, because then everyone's power figures would be different next time he runs their car!

[Evoluzione]

Max_Torque said:

Evoluzione said:

If you want more accurate results over a long interrupted period of time then find a hub engine dyno.

EFA! My engine dynos hold everything to within 1 degreeC at full load for hours at a time, and even then we see significant variation in engine output

Yes i'm quite aware an engine dyno is more accurate, but its use is out of the question for the OP and most engine builders too. In that case the hub dyno produces more accurate repetitiveness than the chassis type.

[spitfire4v8]

stevesingo said:

spitfire4v8 said:

You can do a coastdown on the DD but I'm not sure I know of anyone using it.

The DD modelling to get back flywheel seems very believable for the cars I regularly run and the numbers for nicely modified engines seem to correlate well with what Mignon suggests as the typical maximums for torque/litre as well ..

The argument over wheels figures becomes a moot point though when you consider dynos such as the dynojet give wheels figures around 15 percent higher than the dyno dynamics. Even the guy who invented the dynojet says it over-reads massively.
So .. wheels figures are great, but not all wheels figures are the same

In reality it's a minefield and there's no wonder the internet is packed full of rolling road arguments (I've been involved in many myself over the years), even the manufacturers can't agree on what's being recorded, what hope has the man on the street.
Bottom line I guess is the good advice of using the same dyno to plot the progress of any upgrades, and make sure the operator is consistent in his/her technique. And treat absolute numbers a guide only, comparison between dynos is going to get you in trouble.

there's one hub dyno operator I know who quotes his hub figures as wheels figures, and adds 17/18 percent on top. The industry is often more guilty than the man on the street at exagerating claims, or just not realising what's actually going on with his own testing.

I wouldn't say over reads, they measure differently. Any dyno with a single large roller will have a better tyre/roller interface than an twin roller dyno due to only on contact patch and that when toque is applied through the tyre, it is not trying to climb up the front roller and lose contact with the rear roller.

I friend of mine and I built similar engines, 2.5lt 4cyl 16v. There were two differences in the engine, mine had 0.5:1 more compression and a ported head.

On a DynoJet, measured wheel HP was almost exactly the same 235 v 237 or something and he had much better low down torque. I was initially gutted and perplexed. Then I realised when comparing the graphs that the road speed of the other car was much lower. That car was tested in 3rd and had a shorter diff, much shorter; 3.64 vs my 3.15. My car was hitting 137mph in 4th (which isn't direct), he was hitting 87mph!

When I looked at the results of the coast down test, the losses on both cars were the same at 87mph, only thing was I had another 40mph to go to hit the red line where my losses were double the other car.

When comparing the flywheel power my engine showed significant power advantage above 4k rpm and matched the other engine below that.

So, IMO there is value in coast down tests if only as a comparator of relative losses, especially in twin roller dynos where over enthusiastic strapping down can lose chunks of wheel power which the algorithm can't account for.

This happened on the initial mapping of my car when it made much less than I expected. The dyno operator (whom was not invested in my car nor my type of car) was like, "oh well, not so good then eh!". When I questioned the smell of burning tyre rubber, he was like, "Yeah, sometimes get that. Not on the cars I tune though!"

The DynoDynamics algorithm can't allow for that, so I'm not a fan TBH.

You miss the point, it's not me saying the dynojet overreads, it's the guy who invented it saying that! (or at least the first one built was *calibrated* against a motorbike of the era that wasn't as powerful as the maker said when measured on his dyno, so he changed the reading of the dyno to match.) and from that point on the course was set ..

https://www.hotrod.com/articles/dynojet-chassis-dy...

• reality check .. it's lifted from the internet, so although it certainly seems true from the comparisons I've seen, keep your tinfoil hat on and question everything.

[GregK2]

AW111 said:

Shootout mode has it's origin in the early days of dyno competitions.
Until then, tuners were happy to work with the measured (roller) figures, with correction for atmospheric conditions.

Once people wanted "flywheel" power, and it became a number to brag about, we did a lot of correlation testing and came up with the original 6? classes.

Driveline inertia has an effect in that it takes power to accelerate the gearbox, diff, axles, wheels etc.
You can enter it manually, but of course that means you can use it to "cheat" to get a bigger power figure.
Ditto with ramp rate, correction method, etc.

So the idea of shootout is to lock down most of the variables so you're comparing like to like.
I can look at the shootout chart and know exactly what the variables are.
And that they're set for a n/a twin rotor rotary, with gearbox and tyres to match.

So what I can find looking at other dyno graphs that display the information is that the ramp rate is 150 in 6F vs 100 in 2R
TN is 3.47 for 6F vs 3.18 in 2R

What effect these variables would have on results is what I don't understand, plus as pointed out there are many other potential variables.

I think I just need to run it again at some point.

Edited by GregK2 on Tuesday 13th November 10:21

[AW111]

GregK2 said:

So what I can find looking at other dyno graphs that display the information is that the ramp rate is 150 in 6F vs 100 in 2R
TN is 3.47 for 6F vs 3.18 in 2R

What effect these variables would have on results is what I don't understand, plus as pointed out there are many other potential variables.

I think I just need to run it again at some point.

Edited by GregK2 on Tuesday 13th November 10:21

Ramp rate is the acceleration rate of the run, in kph/sec / 10
So a ramp rate of 100 is 10 kph/sec, so 70-140 kph will take 7 seconds.
Turbo cars prefer faster ramp rates to avoid heat soak. Also, more powerful cars tend to be run over a larger speed range, so a faster ramp rate gives the same run time.

TN is the inertia figure used : it's only about 9% higher for 6F, so won't make much difference.

I think, from your description, that the issues with heat soak etc. that you described are a bigger factor than using the wrong shootout class (except for the slower ramp / longer pull).
There's some difference between classes, but not as much as you think is missing. I'd maybe have run it in 4th gear too, to keep the tyre torque down. What speed does 4th top out at?


OT : I hate dynoing turbo cars.
They're a right pain to keep properly cool, they loose power bigly if they get hot, unless they go bang.
And the big-power dyno queens tend to have a giant turbo with a wicked all-or-nothing power curve that makes life hard for both tyres and the control system.
I've seen one go from 100 to 500 hp in less than 1,000 rpm when the boost came in (then on to 750 odd).

Our old n/a petrol v6 test mule will take 1 run to get everything right up to temp, the next 2 or 3 back-to-back runs will be identical, then the ecu pulls it back a bit because the coolant temp is climbing.
Idle for 5 mins, then repeat.
For similar reasons, one of our development engines is a marinised small-block chev.

[GregK2]

Thanks for the explanation, I really appreciate it.

I believe some of the initial runs were done in 4th, not sure why 3rd was eventually used. 4th is 136mph.

[geeks]

More to the point, if you aren't happy with your dyno operator, I would be finding a new one!

[AW111]

GregK2 said:

Thanks for the explanation, I really appreciate it.

I believe some of the initial runs were done in 4th, not sure why 3rd was eventually used. 4th is 136mph.

That's only 220 kph. Dyno should be good for 250.

Anything with serious power should be run in as tall a gear as possible, IMO.
It reduces the tyre torque, so wheelspin and losses are less of a problem, and it gets the retarder spinning faster with less torque, so it's not working as hard.
This is particularly true for modern turbo cars, where peak torque may be right at the bottom of the rev range.

[stevesingo]

AW111 said:

That's only 220 kph. Dyno should be good for 250.

Anything with serious power should be run in as tall a gear as possible, IMO.
It reduces the tyre torque, so wheelspin and losses are less of a problem, and it gets the retarder spinning faster with less torque, so it's not working as hard.
This is particularly true for modern turbo cars, where peak torque may be right at the bottom of the rev range.

Would wheel speed related losses also increase with a higher gear and therefore higher wheel speed?

[AW111]

stevesingo said:

Would wheel speed related losses also increase with a higher gear and therefore higher wheel speed?

They would, but I'd have to crunch some numbers to see how the tradeoff compares.
For "normal" cars (say sub 250 hp) I'd expect the speed effect to be more prominent, but the tiedown tension and consequent tyre deformation required for high horsepower / torque engines in too low a gear really hurts.

I run my 160 hp MR2 in the 70-150 kph range, but it's barely tied down at all, since it doesn't have the power to spin the wheels in any reasonable gear.

[spitfire4v8]

Hi Adrian I've sent you a PM on a completely different subject

[Stan Weiss]

Dave,
Somewhat off topic. You seem to like to do mathematical modeling. Have you ever taking data logging from the OBDII port a long for the vehicle weight, gearing, tire size, and CDA and calculated wheel HP and Torque?

Stan