Dyno Mystery solved!

[mbohn]

A few related discussion points on dynamometer testing:

1) It’s been said in this forum that dynamometer testing is not repeatable from one facility to the next so comparison of different runs is a waste of time. I ran across the following link:

http://www.dynamometer-info.co.uk/advantages-of-each-type-of-dynamometer.htm

from a fellow in the UK who builds dynos. He does a very nice job of comparing inertial and brake dynamometers and advantages and disadvantages of both as well as many other dyno-related topics. He likes the inertial dyno because it doesn’t require calibration and is very repeatable. When he built his first prototype dyno he compared results with other dynos and says:

In fact I took my own motorcycle to ten different DynoJet dynamometers in the UK over a period of a couple of weeks, and it / they read between 135 and 136 BHP on all of them, as well as on my own dynamometer systems.

This gives me confidence that testing on DynoJet dynos is repeatable and we can safely compare results from one test to another, if they were done on DynoJet dynos by competent operators.

There is a lot of very useful information on the above link. Definitely worth reading if you want to be conversant in dyno talk.

2) In December 2006 I had a dyno run made on my stock ’06 749s and the results seemed really weird to me. I posted the results in this thread:

http://www.ducati.ms/forums/showthread.php?t=15181

(
here’s a blowup of the graph of my data:
http://www.ducati.ms/forums/attachment.php?attachmentid=16760
)

to see if anyone could comment. I didn’t get much feedback but to be honest I felt the data I had been given were not for my bike! Note that there is nothing to identify the bike, owner, date, temperature, baro pressure, humidity, SAE factor, etc, the typical info a DynoJet run produces. Nor was there an altitude correction (pretty important for Denver dyno runs!) The run stops well below my bike’s rev limiter (service manager explained this as being due to the fact that the bike’s tach doesn’t report true rpm !). Although the peak hp is close to what I expected, it happens at least 400 rpm too soon and there are some other problems, more on this below.

Because of that and my desire to map a/f for both cylinders I had Boulder Motorsports, BMS, home of Marty Craggill’s 749R AMA Formula X bike you’ve seen on the podium at VIR and Road Atlanta recently) run my bike again with the emissions probe stuck in the exhaust bungs. There were 3 pulls for each cylinder. And it was all good news.

First, the a/f for both cylinders were quite close to one another, quite consistent across the rpm range and close to the optimum a/f value for max power:

www.ducati.ms/gallery/files/6/7/5/5/AFratiovandhcyls.jpg

I’ve heard on this forum and in other places that Ducati sometimes does not do a very good job of mapping the cylinders. I’ve heard they’re too lean, too rich, both too lean and too rich depending on rpm, have wide variations from horiz to vert cylinders, and all sorts of other complaints. From the data above, I’d say they do a very good job. Or maybe I just got lucky. (I wonder if other folks had the stock setup when they measured a/f?)

Second the power/torque curves make sense and they even compare favorably with the data on the Ducati web site (which are for rwhp):

http://www.ducati.ms/gallery/files/6/7/5/5/hpcurves.jpg

Blue is the run made recently at BMS and yellow is from the Ducati web site (crank hp). It is obvious that the blue and yellow curves have a similar shape although the yellow curve is higher than the blue curve, as is to be expected for rwhp vs crank hp. Note the similar plateau for each curve at around 9000 rpm and the similar shape of both curves near the max hp for both red and yellow curves. For grins, I multiplied the BMS curve by 1.1, green curve, and look how closely it follows the yellow curve. Very close. Peak rwhp from the BMS curve is just under 104 hp, pretty close to what other folks report for this bike. Actually considering I’ve got an SAE correction factor of 1.23 and I’d say it is amazingly close.

The red curve is the suspect hp curve from last December. It is pretty clear to me that my initial suspicion about the red curve is right. I was handed a data sheet from someone else’s bike. If you have any further doubts, note that the red curve is above the yellow curve up to 8500 rpm. How can rwhp be greater than crank hp at any rpm for the same bike? Not possible. Now you can see why I thought that run was weird. Sadly, this work was done at a Ducati dealer. What a great deal for $80.

What all this means is that I’m confident the bike is making as much power as it can with the stock components and I don’t need a fuel injection controller to smooth out the a/f or to get the two cylinder’s a/f closer to one another. I just purchased a Leo Vince SBK Full System Factory Titanium exhaust and hope to have it and a Rapid Bike installed and dyno tuned in a month or so. I’ll report back on what the dyno curves look like then.

3). Bonus question:

Have you ever wondered about the relationship between the rpm at the torque peak and the rpm at the hp peak? Can you explain why the hp peak is always after the torque peak? There is a pretty good physical explanation and a very good mathematical explanation. Either will do. I think it gives some pretty good insight into engine design.

[Pyrate]

Good information, very insightful. I don't know who you worked with at Boulder but Brian Sharp used to work at AMS here in Texas. Great guy and really knows our bikes.

[TAftonomos]

When both dyno's are set up correctly, and maintained correctly, and operated by someone with good working knowledge, they are somewhat comparable. The further you depart from the above, the further the results can be skewed.

You want to really know if something made a difference? Measure transient response (time it takes to go from X rpm to Y rpm). If your parts/improvements helped, it should drop.

Easiest/cheapest way would be going to the dragstrip. I've seen cars that dyno'd 400+ HP trap higher than cars that dyno'd 500+ hp, similar weights. Trap speed is a function of HP, ET is a function of traction and launch technique.

[Colt45]

HP is (torque x RPM)/correction constant. HP rises with RPM and will rise faster than the torque falls off in most cases. Peak torque is at peak volumetric (cylinder filling) efficiency. Above that volumetric efficiency falls off, but normally not as fast as RPM rises, thus the HP peaks later.

I think your A/F curves are abysmal! The vertical cylinder is seriously lean around 10k and gets just rich enough briefly at 7500. The Horizontal cylinder is just the opposite.

The WOT curves should be just about dead flat and between 13.2 to 13.6 from 4k to the rev limit, depending on what ratio your engine likes.

[Pinocchio]

Originally Posted by Colt45

I think your A/F curves are abysmal! The vertical cylinder is seriously lean around 10k and gets just rich enough briefly at 7500. The Horizontal cylinder is just the opposite.

The WOT curves should be just about dead flat and between 13.2 to 13.6 from 4k to the rev limit, depending on what ratio your engine likes.

+1 on the air fuel ratio...it should be a lot smoother for sure! and they should almost overlap one another.

[OldBaldy]

Typical poor stock fueling across the cylinders, The vert fuel is seriously lean, as normal, and the bike woud be smoother and make more power if that cylinder was running round 13.0-13.2:1 AF on pump gas.

As for the dyno types link, well, suffice it to say that anyone building a fuel or spark map MUST run in brake mode - to do a decent job. It absolutely requires steady-state RPM for a number of seconds to log a good average lambda/AF value for that RPM/TPS intersect - and preferably on a dyno that has the capability to DETERMINE what air-fuel ratio is best at that specific intersect, something most inertia dynos cannot do, even in brake mode.

For simple and quick approximations of power across a rev range, the inertia dyno is fine, but a good brake dyno is MUCH more useful for actual tuning, IMO.

Of course, any inertia dyno is influenced by the rotating mass of the gear-train and wheel/tire assembly, so unless the dyno operator is able to know and input the actual rotating mass of any specific bike (he's not as it is only assumed to be the same for every bike!), there will ALWAYS be some variances indicated even with identical motors or motors that actually produce the same power.

The example sited in the link about a drag bike running through detonation at 6000RPM may be valid in a drag race - but how would one feel if you did not know about the detonation happening at a specific steady-state RPM while running out on the open road in the middle of nowhere? Not so good, then....

Regardless of all this, any dyno that is repeatable, is a useful tuning tool, and certainly better than NOT having one to check your tuning. (I'd be thrilled with an inertia dyno with brake mode in my basement! )

[TomTom]

I hope you put the wrong AF chart up or else you got it all wrong.
And for 80usd I wouldn't expect much of a tuning, nothing strange there.

Your dyno building friend is hardly taking the full dyno environment into the picture, just states that the same roller always takes the same time to accellerate.

Of course it does, but other than that you have air supply, air temp, size of dyno room, total ventilation of dyno room... I can go on..., just so many parameters it's impossible to have the same testing environment.

Not even on the same dyno with the same operator you can have a stable environment, unless it's say an F1 development dyno.

However, used the right way the dyno is a good tool for mapping.

[Synergy]

Originally Posted by mbohn

What all this means is that I’m confident the bike is making as much power as it can with the stock components and I don’t need a fuel injection controller to smooth out the a/f or to get the two cylinder’s a/f closer to one another. I just purchased a Leo Vince SBK Full System Factory Titanium exhaust and hope to have it and a Rapid Bike installed and dyno tuned in a month or so. I’ll report back on what the dyno curves look like then.

Like others have said, your A/F could use some work, once you bolt on the Vince FS most definitely will need some FI tuning. If your getting this in depth on the technicals for your bike. I would suggest, installing the FS, using a PowerCommander or such and getting the cams degreed. This will increase you performance quite a bit over stock and the motor will be running at optimum.

[mbohn]

Originally Posted by Colt45

I think your A/F curves are abysmal! The vertical cylinder is seriously lean around 10k and gets just rich enough briefly at 7500. The Horizontal cylinder is just the opposite.

I've not seen data on a stock bike with both cylinders mapped or for that matter a bike that has been tuned correctly. Sounds like you have so how about posting some a/f curves? Also, have you seen a stock bike that is much better?

[mbohn]

Originally Posted by Synergy

Like others have said, your A/F could use some work, once you bolt on the Vince FS most definitely will need some FI tuning. If your getting this in depth on the technicals for your bike. I would suggest, installing the FS, using a PowerCommander or such and getting the cams degreed. This will increase you performance quite a bit over stock and the motor will be running at optimum.

The plan is to install the Leo Vince with a Rapid Bike controller then dyno tuning. Should be fun!