Horsepower claims, Top speed and Accuracy of Dynamometers

[StewartD]

220px-Dynamometer01CJC.svg.jpg

400px-Dyno_schematic.svg.jpg

The issue of Dynamometer accuracy came to be discussed in the thread 'Wanted Ducati 250 Clutch. Also under discussion was accuracy of Claimed horsepower figures and rear wheel horse power. I think a new thread is justified so I will kick it off.

I think that dynamometer horsepower figures would vary more due to test conditions than variation between dynamometers. By test conditions I mean the humidity and air temperature. The diagrams above and the following text is from Wikipedia:

"A dynamometer consists of an absorption (or absorber/driver) unit, and usually includes a means for measuring torque and rotational speed. An absorption unit consists of some type of rotor in a housing. The rotor is coupled to the engine or other equipment under test and is free to rotate at whatever speed is required for the test. Some means is provided to develop a braking torque between the rotor and housing of the dynamometer. The means for developing torque can be frictional, hydraulic, electromagnetic, or otherwise, according to the type of absorption/driver unit.
One means for measuring torque is to mount the dynamometer housing so that it is free to turn except as restrained by a torque arm. The housing can be made free to rotate by using trunnions connected to each end of the housing to support it in pedestal-mounted trunnion bearings. The torque arm is connected to the dyno housing and a weighing scale is positioned so that it measures the force exerted by the dyno housing in attempting to rotate. The torque is the force indicated by the scales multiplied by the length of the torque arm measured from the center of the dynamometer. A load cell transducer can be substituted for the scales in order to provide an electrical signal that is proportional to torque.
Another means to measure torque is to connect the engine to the dynamometer through a torque sensing coupling or torque transducer. A torque transducer provides an electrical signal that is proportional to the torque.
With electrical absorption units, it is possible to determine torque by measuring the current drawn (or generated) by the absorber/driver. This is generally a less accurate method and not much practiced in modern times, but it may be adequate for some purposes.
When torque and speed signals are available, test data can be transmitted to a data acquisition system rather than being recorded manually. Speed and torque signals can also be recorded by a chart recorder or plotter."

Note that it is not important what is absorbing the power, "...it can be frictional, hydraulic, electromagnetic, or otherwise, according to the type of absorption/driver unit." What is important is that the measurement is done by a load cell transducer or by scales. Both methods should give a force reading accurate to about 1%. This force multiplied by the distance the load cell or scale is from the stator centreline will give torque.
i.e. Force (in Newtons) x Distance (in metres) = newton.metre (Nm)

The r.p.m. is measured concurrently and a similar accuracy is quite feasible with modern equipment. If the rotation speed is expressed in S.I. units of radians/second then Torque x Rotation speed = power; N.m x rad/s = Watts. (A radian expressed in degrees is 360/pi)

Cheers,

Stewart D

[DanM]

[quote="StewartD I think that dynamometer horsepower figures would vary more due to test conditions than variation between dynamometers. By test conditions I mean the humidity and air temperature.
Stewart D[/quote]

And altitude.

[StewartD]

Dan,

Yes, good point about altitude.

All,

Sorry guys, I didn't give the new thread a very appropriate heading. I wanted to continue on with the subject that Rocla, Bob had been discussing in the thread 'Wanted Ducati 250 clutch', That is the claims manufacturers make for various motorcycles, the actual horsepower measured at rear wheel on dynamometers and the real world top speed of these machines. The accuracy of the testing and where errors come in is interesting as well.

About altitude, humidity and temperature:

I believe the Society of Automotive Engineers (S.A.E.) have standard conditions for dynamometer testing and correction factors that are factored into the results of dynamometer tests. These correction factors should ensure all test results are comparable. Does anyone know about these and whether they are commonly used by motorcycle industry testers?

Cheers,

Stewart D

[Paddy]

Ahhh Dynos;s
I have had quite a bit of experience with dynos. They do vary quite a bit. On a little Ducati i dont know just how much that would matter. 10% on my 200TS might be 1bhp but on my audi would be 30bhp. even tyre pressures would be a factor on the TS.
The real value of a dyno is as a comparator. run, fiddle run fiddle run will give you accurate results for your fiddling. A before and after is a better use than a ...this should produce 15bhp, what's it actually producing ? test.
As for top speed ? well i would guess on a little ducati, engines could vary on bhp as much as 10% just on manufacturing tolerances. for example i remember years ago running a std MGB and one day dialling in the cam timing sprockets with a degree disc. I didn't dyno it but the difference was very noticeable on the road. that was basically just manufacturer's tolerances.
You would have to reckon on USA petrol at 91 octane being a factor over UK petrol at 99 octane. even the weight of oil will be a factor in a little engine. On my audi we found that std unleaded at 97 octane gave a 10% drop in top BHP over super unleaded at 99 octane.
I would say that you simply cant compare one dyno's reading to another but you can compare readings from the same dyno.
HTH

[Stan Lipert]

Three singles that I've tuned on a Dynojet 250i:



Notes:
Points and ignition inductive pickup would not get stable readings, therefore runs made by MPH, in ? gear.
Sebring stock
Mk3 w/ meg
250 is wide case scrambler, Mikuni carb and open exhaust.

All modern dynos that you can put a bike on have compensation available for temp, and air density. Some compensate better than others.
All engines make more power with cold, dense air if fueling is correct.

Lower octane gasoline will make more power than higher octane gas in a bike if tuned correctly, unless there is detonation.

Stan

[StewartD]

Paddy,
The improvement in output from getting the cam timing spot on is something that a Barry Jones (of Melbourne, not the U.K. Barry Jones!) advocated for all Ducatis. Barry had Italian Motorcycle Engineering in Melbourne. He ran dyno sessions for the Ducati Owners Club of Victoria in the 1990s. I will try to get some of the results.

He had 'dialed in' the cam timing of my 600TL Pantah before I owned it. He had to make up stepped keys for the belt pulleys to achieve the factory timing. The difference between it and a standard 600 Pantah was remarkable. There was bags of extra torque right through the rev range.

Stan,

The curve for the 250 Mark 3 peaks at 21.36 hp, very close to what the 'Cycle World' test achieved (22.2 hp and 156kph/97 mph), mentioned toward the end of the topic 'Wanted Ducati 250 Clutch'.

Apart from megaphone, was this Mark 3 standard? The Mark 3 were very similar to the Mach 1 I believe. Did the owner ever try a top speed run? Do you know what size carburettors the 3 machines had fitted? How was the Air/fuel measured? Is the dashed red line the optimum ratio?

I do not understand the note regarding 'Points and ignition inductive pickup.....'

I assume the Mark 3 was in third gear, with tall final drive ratio, and the scrambler was in top gear with much lower final drive ratio.

Cheers

Stewart D

[Bevel bob]

Most interesting, I really must get my Mark 3 on the dyno. The narrow case mark 3 was in a very similar tune and equiped as a Mach 1,the widecase differs and did not achieve the same performance .It would have been interesting to have rpms and carb information.

[Stan Lipert]

Stewart,
The narrow case Mk3 had the stock SSI 30 with no porting, but a fresh rebuild. The red line is optimum, but it depends on where the wide band o2 sensor tube pickup is located. This would explain why optimum air fuel ratios are different for the three bikes.
The dyno uses an inductive wire pickup to count the sparks. I can't show RPM or torque without counting sparks.

The scrambler was converted to a Cafe, who knows what gearing?
Each bike ran up to around 9K RPM, one may guess a comparison bu looking at the curves.

Speculating on top speed is just that- speculating. There are MANY variables for top speed.

Stan

[Bevel bob]

The Mark3 would appear to be pretty well set up, maybe a little rich on the slide or needle ,any idea what the main jet was?, I think the ssi29d was the standard carb. With the unported head Mach1 power tailed off at 9000 and the points were getting distressed too. My current Mark3 has been ported and with better ignition shoots up to 10000 and beyond, , I would prefer that it did not!!.Don't think the motor will last at those revs. I would be happier with the reliability of a standard head. I think the Mach1 produced about 24/25bhp.And could be coaxed too around 100mph if in top tune. I bet Lacey has some interesting graphs!.

[machten]

The improvement in output from getting the cam timing spot on is something that a Barry Jones (of Melbourne, not the U.K. Barry Jones!) advocated for all Ducatis. Barry had Italian Motorcycle Engineering in Melbourne. He ran dyno sessions for the Ducati Owners Club of Victoria in the 1990s. I will try to get some of the results.

I have no doubt about this. I've dialed cams on a number of 750 GTs and Sports. Ducati specified the cams as being +/- 5 degrees of accuracy. I've dialed twin's cams as being "out of the factory" as -5 and +6. The effect of fixing that using offset woodruff cam keys, you wouldn't believe. It makes a hell of difference. An exhaust that rumbles beforehand, finishes "barking".

I have dialed both my 750 Sport and my 750 GT using offset cam woodruff keys. The difference is exhilerating.

I suspect there is a greater benefit for twins in doing this, but I will at some stage get to doing the same test for my singles.

Kev