I had some interesting experiences with rolling road dynos recently and I'd like to share them and try to explain why...Part1 I recently dyno'd my car on a DynoDynamics rolling road with DOS based software. FWIW the RWHP figure was 215hp. The acceleration rate of the engine was 550rpm/sec. During the running both I and the operator could smell burning rubber from time to time. The operator didn't seem too concerned and when questioned he said it was tied down correctly. It was tied down over the rear arms which in turn compressed the suspension adding both camber and toe. A lot of camber and toe equals scrubbing and smelling rubber. The upshot of the run was that I thought this was a little low and left me with doubt in my mind. Is the engine bad or are the rollers bad? I spoke to some fellow owners and it became apparent one other person using DynoDynamics have run it differently. DynoDynamics rollers have shootout modes that locks down the settings so they can't be fudged, but the earlier DOS models can be changed. My run was in shootout 4R and the engine accelerated at 500-500rpm/sec. The other persons run was done out of shoot out mode at 100-110rpm/sec. So what? In order to accelerate the drive train and the engine it's self, it takes power. So, providing the inertia of the drive train is the same it will take more power to accelerate that same mass 5x faster. The inertia of the drive train can be input in to the software, and can be determined by the operator. But, I know it wasn't done for my car. Anyway... The cheapest way to find out is the engine bad or are the rollers are bad would be to go to a different set of rollers for a power run. As it happens PBC M3 had run his similar engine on a set of rollers earlier in the year about 90miles from me, so I booked in. What is also convenient is PBC M3's engine is of very similar spec to mine... stevesingo 2.5 11.9:1 compression 29.5mm head with 38.5mm Intake and 32.5mm Exhaust valves. 292/284 Schrick cams-106/106 timing 17cm Runner Airbox 3.15:1 diff in 4th gear or 59.12rpm/mph PBC M3 (correct me if I'm wrong) 2.5 11.9+:1 compression 215 head and valves. 292/284 Schrick cams-102/106 timing 17cm Runner Airbox 3.91:1 diff in 3rd gear. (I think) or 97.88rpm/mph This dyno was a DynoJet, So I will explain the differences. DynoDynamics: 2 small interconnected rollers in which your tyre sits. Your car is strapped down to ensure equal contact on both rollers. Essentially 2 TWO tyre contact patches per tyre. Load applied by eddy current retarders, which effectively brake the rollers and the torque generated is multiplied by engine speed to give the power. The advantage of this is you can specify a fixed acceleration rate, say 150rpm/sec. Why is this an advantage? If you increase power of your car the acceleration remains the same for the test, therefore negating the effect of accelerating the mass of the drive train faster due to more power and therefore losing measured power. Flywheel power is determined by adding a fixed transmission loss factor which is related to the shootout mode used 4R being 4cylinder Rear drive. DynoJet: Single large roller/drum on top of which your driven wheels sit. Single contact patch. The car is strapped down in order to stop it going both forwards and backwards as it could go either way. Although some DynoJet rollers have an eddy current retarder same as the DynoDynamics, the power measurement is made without a load being applied by the retarder and relys on the inertia of the roller for the load. (the operator told me the drum weighs 1000kg) The drum has a fixed inertia and it takes a fixed amount of power to accelerate the drum at a fixed rate. The disadvantage is that if you have more power not will you accelerate the drum faster, but also the drive train faster and power spent accelerating the drive train is not recorded. Flywheel power is determined by allowing the dyno to coast down. The theory being the fixed inertia of the drum will take a fixed time to slow from a fixed speed. If your cars wheels sit on the drum, the drum will slow faster therefore that drag is measured and from that the hp that same drag will absorb during acceleration. So, here is the result of the DynoJet run. So where did the 25RWHP come from? Well from my ECU datalogging, the engine rpm rate was and average of 360rpm/sec, so this will have had a small effect, but it is also worth remembering the engine will accelerate faster at peak torque and slower at lower torque areas of the ref range. My guess is the two contact patches and the tight strapping down played a significant part. What about the flywheel power? Well, the DynoDynamics just adds 18.5% to the wheel power, so if there is excessive drag, it is not measured. The DynoJet at least tries to measure the drag, even if this is not truly representatie asit doesn't take in to account how efficient the drive train is at transmitting torque. As an aside, looking at the raw data, the losses seem to be proportional to road wheel speed only, but I'll have to plot a graph to investigate more. So, what have I learnt? Well, nothing I didn't already know, nor suspect. I will leave you to make your own conclusions as I don't want to get in to a slanging match about which is best, both have their drawbacks and advantages, I am attempting to highlight them for all. But more in part 2...
Part 2... So how did the comparison of similar engines go on the same dyno? Taking the raw data I plotted a graph for RWHP... So, I should be disappointed that as I have the same displacement, cams, airbox and a 29.5mm big valve head, I can only just match PBC M3. It is not quite as clear cut as that... Flywheel HP How the F*** did that happen? You all remember in the previous post I stated that the drive train losses are almost directly proportional to road wheel speed, and if you look carefully at the specs you might have noticed the different diff ratios with the corresponding rpm/mph. Working again from the raw data I can see a comparison of losses 76.6mph PBC M3 17.26hp losses Steve 13.7hp losses But here is the rub... at 76.6mph PBC M3 is producing more torque at 168lb/ft to my 151lb/ft. This might explain some of the extra 4hp losses, but not all and certainly not the 17hp difference in total losses. What is more pertinent is that due to the gearing PBC M3 is at 7500rpm at 76.6mph, my engine is at about 4150rpm. By the time I hit 7500rpm my road wheel speed is 133.6mph with 35.5hp losses. It is clear that the road speed has an effect on measured power and this seems reasonable and with the tyre being deformed at the contact patch at a higher rate and the tyre scrub due to toe having a greater effect as speed rises. The interesting thing is the DynoJet's coast down to determine flywheel power from the drag is giving a result that both PBC M3 and I concur is reasonable given the differences in engine spec. My 2nd gear takes me to 75.5mph at 8000rpm as opposed to PBC M3's 3rd giving 82.4mph at 800rpm. This may have given us a more accurate comparison, but I didn't think of it at the time. Learning points... Wheel figures aren't the be all and end all as there are factors that have an effect. When comparing wheel figures, make sure it is like for like. All pretty interesting really and my conclusion aside from taking the engine out of the cat and putting it on an engine dyno at a fixed acceleration ramp rate, you will never know. The best option is probably a hub dyno reading hub figures only. This would remover the tyre drag issue, but ramp rates will need to be consistent. Comments welcome...
Hi, Getting the rear wheels as high up on the rollers as possible on any Dyno will always bring higher figures. The unwritten rule always has been- When selling a Race Car, Dyno very early in morning, over inflate the rear tyres if RWD, strap loose and get the numbers in 2 runs. When buying a Race Car Dyno mid afternoon, reduce rear tyre pressure if RWD and strap it down so tight that the rear wheels can not 'climb' up onto the top of the Rollers and take about 4-5 runs to get the numbers. You can alter a E46 M3 by around 30 bhp pending where the rear wheels sit on the rollers and having the fan blowing in slightly the 'wrong' direction. Try telling that to all those that have purchased Tuning Maps and gained 30 BHP plus from just a tuning file or remap !!!! Reagrds, The Gorilla.
Are you sure about those ramp rates? 100rpm/s is very low and a power run would take some 40seconds if a 4000rpm spread was done the ramp rate and drivetrain inertia correction factor (different to coastdown losses) are governed by the shootout mode selected i.e shootout 4, shootout 4F etc this is the same with DOS and windows though preset factory numbers may have changed between operating systems. its only non shootout where these can be manipulated the same way that you can ad any correction you want in a non shootout mode. for example 10-20km/h per second are common ramp rates for some of the shootout modes as for strapdown yes it is critical when doing comparisons and consistency as long as you know this there are no real issues until you make the mistake of comparing another persons graph
I can only go off what I was told by the other guy. Regarding comparisons, if I buy two measuring devices, say a torque wrench, from two different manufactures it shouldn't be unreasonable to expect them to give similar results. In this regard rolling road dynos are somewhat flawed.
Hi, Quote- '' as for strapdown yes it is critical when doing comparisons and consistency as long as you know this there are no real issues.....'' Strap down is one thing, but differeing wheel tyre sizes altering rolling circumfrance changes the dynamics even more when people make comparissions. Its nearly always overlooked. Regards, The Gorilla.
you have to remember what you are measuring, different brands of dynos are often not even measuring the same thing.
i always went overboard with doing baselines if i wanted to do a decent job of a comparison, so if i change tyres and wheels i reestablished the baseline and went from there. This way if the absolute values changes then im still able to track what is going on. you have to know what you are measuring and variables involved and minimises change to variable not of interest
Hi, Quote- ''you have to remember what you are measuring, different brands of dynos are often not even measuring the same thing.'' Exactly, given the variables it often makes comparissions somewhat irrelevent. Regards, The Gorilla.
i was more meaning that an inertia dyno and eddy current dyno are not measuring the same thing (even when influential variables that are not of interest are controlled). one is measuring hp based on accelerating a fixed inertia load, the other it is measuring power when doing work against a changing load. the way the engine accelerates between the two is different so the outputs will be different in the same way a transient output will differ to steady state.
