Rolling Road Testing Information and facts

The information below was initially put together so that people can understand the rolling road procedure in a little more detail, and try to make it clear about the errors that can creep in when performance testing cars on a rolling road. We tried to keep this report on track to be just about Dyno Dynamics, however it has gone into more technical detail to get various points across.
Hopefully this information will give people more of an insight into the Technical side of the rolling road testing procedure along with things to be aware of and look out for.
Please note that the information on this page is very much work in progress and we will add other sections as and when we get time to write them.
Last updated by Keith at JKM Performance on July 22nd 2009

Technical Introduction

JKM are well aware that any correctly operated Dyno Dynamics Rolling road system regardless of location or operator does not hand out ‘Inflated estimated flywheel numbers’ as some other Rolling roads in the country frequently do.
However there appears to be a belief that the JKM Dyno Dynamics rolling road somehow reads lower than other Dyno Dynamics rolling roads in the country?
We have been asked this a few times now, and taking into account the time it takes to answer everyone individually - it is obviously a lot easier to write this once and direct people to read this page.
I am sure that this write up will not change the views of everyone, especially those with big numbers that they know they are unlikely to see at JKM - these people may actually question their 'better than expected' result themselves but not publicly admit to it.

JKM are happy to lay out the facts about performance testing on a Dyno Dynamics rolling road - please note that not all of the information below is relevant to other makes and models of chassis dyno as JKM only have in depth working experience of the Dyno Dynamics system. Having said this some of the information below is relevant such as inputting the correct atmospheric conditions that the testing is being performed under.
JKM do not perform any form of favouritism to any tuning company or software provider to promote one type of product over another, if a competitor’s software performs better it will reflect in the results, Likewise if it performs worst then this will also show in the results.
With our own ECU mapping, again we always show true results, it is better to know the true power than be fooled into thinking you have more than you actually do. Some of our customers may have less power ‘on paper’ than the car next to him, and yet they are considerably faster at track days and various timed events.
I am not sure how many other tuning companies would be so open about the ‘facts’ with their rolling road testing – What’s more we strongly believe that some other rolling road operators in the UK are either unaware of the fact their incorrect operating methods are creating higher numbers, or conveniently they are choosing to ignore them as it may reduce the numbers at the end of the test !

The belief by many people unaware of the factors laid out in this report is that because the car is run on a Dyno Dynamics rolling road the result should be consistent regardless of location up or down the country? This would be a valid statement if the same operator and cooling arrangements where used in each case. However the truth of the matter is that the results that you see on any Dyno are affected by a large number of variables; many variables are taken care of by the software (if it is told the truth) however a huge influence is the actual rolling road operator and the choices and actions that they make when running the car.

Just as any tool is only as good as the operator – this is true with the rolling road as well. It is a tool that must be used correctly or else you will get either inconstant – or incorrect readings.

If the rolling road is used correctly, then you will get consistent dyno readings regardless of the time of year or temperature etc. This is assuming that the cars engine is running in the same mechanical condition for example it is running on the same fuel grade and hasn’t developed a running fault that results in it genuinely being down on power.
On first receiving and using our rolling road JKM made a conscience effort to try and learn all of the errors that can creep in, learning how you can trick the system – so that we ensure that these situations do not happen. The end result is that the customer gets accurate results that are repeatable.
In order to explain matters further it is important to understand what is actually happening when your car goes onto a rolling road and what the operator is entering into the data aquistion computer.

Car strapping down on the Dyno.

Once your car has been driven on to the rolling road dyno bed, it is driven at a slow speed so that it is ‘squared up’ relative to rolling road drums– on starting the actual run(s) once the car is strapped down we then switch off any traction control or ESP system so that we are able to drive the car under full load with only 2 wheels spinning and no traction control working against us.
Strapping the car down correctly is extremely important for more than the obvious reason of keeping the car on the dyno bed.
The car must be strapped down to match the ball park power level that is going to be made – for instance a 600BHP car is strapped down considerably different to a 100BHP car. However in both cases the car must be strapped to ensure that it remains in contact with both dyno drums throughout the testing – put another way there must be 4 four contact patches onto the dyno bed during the testing.

We will also change our strapping methods if the vehicle is either wheel spinning due to poor quality tyres or producing a torque level that is extremely high at low revs.
The Dyno Dynamics software does not measure the speed of both dyno bed drums and it makes the ‘assumption’ that the operator has correctly strapped the car down in such a way that both tyres are remaining in contact with both drums throughout the test.
There are to be 4 four contact patches with the dyno drums during the test. The front drum of the dyno has a knurled machined finish which gives the tyres fantastic traction to ensure that you do not get wheel spin during the power testing and an inaccurate result.
The back drum is an idler that forms an important part of the dyno’s mechanical inertia, but importantly it must remain ‘in contact’ with the cars tyres during testing.
Below are two examples of cars that have both been run 'Correctly' as per the correct Dyno Dynamics strapping method – one is a Skoda Octavia TFSI (Manual) and one is an Edition 30 Golf (DSG).
Both of these cars were then run ‘Incorrectly’ by releasing the tension on the straps slightly so that the cars where still under control on the dyno bed - but in both cases the vehicle was able to ‘climb’ up onto the front knurled drum of the dyno. Nothing else has been changed during the testing and these tests were performed less than a minute apart. To conclude the car was then re-strapped correctly and the results became as per the first run again! No surprises there then. For another example on a Leon Cupra TFSI - see this link.


K04 Turbo Powered Skoda Mk2 Octavia VRS TFSI (Manual) Click on the image to Enlarge	Volkswagen Golf Edition 30 (DSG) Click on the image to Enlarge

To explain why this happens, in allowing the car to climb up higher on the front roller this action frees the load off of the tyres on the back idler drum – but the dyno ‘believes’ that this back drum is still being driven by the car. The "mathematics" as to why the power increases is due to the fact that the Rear Rollers (idlers) are accounted for in the Dyno ‘Mechanical Inertia’ value, so if your not turning the rear drums you are getting something for nothing, how much depends on the Ramp Rate (Acceleration Rate) of the Dyno at the time, basically the faster the Ramp Rate the more you get.
It must be kept in mind that once you go up in power level on high power rear wheel drive cars - such as 500HP upwards, these cars definitely require substantial downwards force for traction, but we would certainly let them "climb" forward a little, as you would introduce wheel spin if you didn't allow this to happen.

Ensuring that cars are strapped 'tight' gives better repeatable results, if the car has enough torque, it will climb onto the front roller, the idea is that when it is strapped tight, and the car tries to climb the roller, it "bites" into the roller, that way there is no wheel spin so results are repeatable.
As a rough guide for owners to be aware of what is considered ‘normal’ or ‘not normal’ if the car moves an inch or so forward, this is considered normal, but if it moves 4-5inches forward and upwards on the drum so that you can easily see under the back of the tyre it is considered a bit excessive and a better strap down method should be used - this decision should be made by the dyno operator or an experienced dyno assistant if present. By the same point, if the vehicles tyres look as though they are about to burst from the tension being pulled down on them, then that is also a bit excessive as well and will most likely be absorbing excessive power.

While some people will be critical of this potential variable, (an experienced dyno operator will always strap a car correctly and on it's own merits to produce repeatable, consistent results), there are far more variables to consider that will effect power readings such as the effect of the Atmospheric conditions of the day.
To conclude, the result of testing with incorrectly straped vehicles (which is a method that we know some companies in the UK with Dyno Dynamics rolling roads are using, as we have seen it ourselves) is that the torque and power will increase – There are not any hard and fast rules as to how much the torque and power will increase, however as seen from the graphs above, a level of between 20BHP to 30BHP on forced induction 4 cylinder engines producing in excess of 300BHP is quite common. We dont make a habbit of testing in this way though for the very reasons outlined above!

Atmospheric conditions & Standardised power figures.

Unfortunately rolling road testing is not as simple strapping the car down correctly onto the dyno bed and away you go… It is full of ‘Atmospheric corrections’ and processes to standardise the measured figures to ensure comparable results under different atmospheric conditions.
Every engine is effected by the atmospheric conditions of the day and so the line of thought that people have that the conditions of the day are sometimes better for dyno testing is true – However due to the environment changing from day to day , excluding factors such as fuel grades in use then the results need to be standardised in some way. Automotive Engineers have calculated methods to standardise the power measured to become what is commonly known as a ‘corrected’ power measurements. JKM always state corrected power measurements on result sheets unless requested by the customer.

On the Dyno Dynamics rolling road as per most other rolling roads, the effect of the local environment builds up what is called an 'Atmospheric Correction Factor’ which is a calculated number typically somewhere between 0.9 and 1.1 that the ‘uncorrected’ power of the day is corrected by.

This can be seen on the Dyno Dynamics screen in two places, as shown in the picture to the left.

One location that shows this correction factor is on the screen where the conditions of the day are being entered into the dyno, by the operator.

The second location is on what is called the 'Bar screen' where the Atmospheric correction number again appears.

For an overall example of how the Atmospheric Correction factor works, take this situation…
An engine is rated at 90BHP and it is put onto the rolling road on a winter’s day and the engine makes an ‘uncorrected’ power of 95BHP. This could be due to the cold air temperature of the day and the high air pressure– however as we are using the rolling road with the Atmospheric correction on - the Atmospheric correction factor could realistically be something like 0.948 and applying this would make the power measurement become 90BHP as the standardised power............(95BHP x 0.948 = 90 BHP)

Bring this same car back on a hot day in the summer and the engine could be making an uncorrected power of 85BHP – however we are again using atmospheric correction and this could be something like 1.0589 due to potentially low air pressure on the day and the hot air temperatures. All other factors being equal such as fuel grade etc, by applying the atmospheric correction factor this will again make the power measurement 90BHP as the standardised power............ (85BHP x 1.059 = 90 BHP)

You can see the various atmospheric conditions during testing at the bottom of the Dyno Dynamics shoot out results graph, however the overall correction factor is not shown. There are lots of combinations of atmospheric conditions that can create the same correction factor.

It is well worth knowing that there are several different methods of atmospheric correction in use across different rolling roads and rolling road manufactures.
In certain situations specific correction factors are applied to individual countries for mainly historical reasons that the country has always used type x correction factor.
As an example of the various correction types that are officially listed, power correction could be in accordance with the following:
• SAE-J1349 A widely accepted well working correction in use by lots of dyno/rolling road manufactures as default.
• DIN 70020 A popular correction in use; however in a lot of conditions it will give a marginally higher result than others.
• EEC 80/1269
• ISO 1585
The difference between the most generous to least generous can vary by approximately 4%

This alone obviously doesn’t help when customers choose to compare results across different makes of dyno or try to blame dyno X for being higher than dyno Y or vice versa. It is not necessarily the rolling road or the operators fault under these situations – it may be purely down to the way their rolling road works things out in the background combined with their dyno’s transmission loss calculations if they are stating an estimated flywheel figure (Transmission loss is a different subject all together and one that we will cover at a later date)

Atmospheric Inputs that Influence engine power

Ambient air temperature.... (AT)
The temperature of the day has an important influence on the power an engine can produce. It is well known that on hot days an engine will produce less power when compared with colder days, this is due to the hot air being less densely ‘packed’ with oxygen when compared with a cold day – as a side note this is the reason for an intercooler on turbochaged cars, where the air has been compresed and heated up, the Intercooler assists in the recovery of the 'charge' density by cooling the air back down.
Cold dense air means a greater mass of oxygen per power cycle and hence more power can be generated. If a dyno operator enters the Ambient Air Temperature of the dyno cell or room incorrectly into the Atmospheric correction screen, this can easily have the effect of tricking the system and effecting the power measurement.
An example of this is if the dyno cell/room temperature where reported as being considerabily hotter than it actually is, this will increase the dyno correction factor and result in an increase in the torque and bhp reported.

Barometric pressure..... (BP)
On days where the Barometric Pressure is 'High' – this has the natural effect of making an engines cylinders ‘fill’ with air easier when compared with a day where the pressure were lower.
Basic physics dictate that cylinders on an engine will find it easier to fill with air when the barometric pressure of the day is for example 1025mb - compared to a day when the air pressure were lower (for example 980mb).
Putting it in basic terms 'a cylinder that is filled with more air will be able to produce more power due to the increase in oxygen content present'.
It is very easy to simply enter in an incorrect Barometric pressure into the Atmospheric correction screen, Falsely telling the dyno that the barometric pressure is lower than it actually is on the day. This has the effect of making the correction factor increase for reasons explained above and the power and torque curves will increase across the curve.
The dyno operator can easily ‘influence the result’ in this area. I have seen this ‘applied on purpose’ by some dyno operators as I have checked the Barometric Pressure on the day and know that what is reported on the dyno result screen is incorrect.
This really is an instant power increase for a few taps of the keyboard.

Relative Humidity..... (RH)
The maximum amount of water vapour (Humidty) in the air is dependant on the temperature and the Barometric pressure. High Temperature and low Barometric pressure make the percentage of water mass (Humidity) in the air increase.
The effect is that each molecule of water vapour displaces one molecule of oxygen available and reduces the density. The maximum power that an engine can produce is limited by the amount of oxygen taken into the cylinders.
Therefore, this should be taken into account in the atmospheric correction. A higher relative humidity will increase the atmospheric correction; where as a lower relative humidity will reduce the correction. In the real world humidity makes a ‘very small’ difference to the power measurements and it is often ignored by dyno manufacturers. Dyno Dynamics choose to include this input.

Intake Air Temperature..... (IT)
The one controversial input on most rolling roads is the intake air temperature probe. It is the cause of much discussion and debate amongst rolling road operators and certainly can have a significant influence on the power measured – more so if the air temperature measurement is taken from a questionable location (such as being influenced by radiated heat from exhaust headers).

Shown in the diagram to the right is the popular SAEJ1349 atmospheric correction factor.

cf = is the actual dyno correction factor
Pd = the pressure of the dry air, millibar - the pressure of the dry air Pd is found by subtracting the vapor pressure Pv from the actual air pressure.
Tc = ambient temperature, deg C

It is important to note that no intake air temperature measurement is used in the above SAEJ1349 power equation; instead only the ambient air temperature is used. It is agreed by most experienced dyno operators (certainly amongst experienced Dyno Dynamics operators ) that the intake air probe must only be used to reflect the actual intake air temperature that the engine bay is taking in which will be close to the ambient air temperature of the day (typically within a maximum of 10-12 deg centigrade).
It must not reflect an inlet manifold temperature (as per the engines air temperature sensor if fitted) and the air probe should not be allowed to influence the power measurement result.

By using an Inlet Air Temp Probe in some way to ‘influence’ the Atmospheric correction, this allows the dyno operator to potentially ‘influence’ the measured results, either accidentally or on purpose.
It is important to know that if measuring the intake air temperature and using this to influence the atmospheric correction factor then as the air temperature increases, so to will the dyno correction factor which will ultimately increase the torque and power level.

As an example taking a turbocharged car where the standard intercooler is not sufficient to keep the inlet air temperatures low........

Given the above situation the actual inlet air temperature measured is likely to get significantly hotter than the days ambient temperature value. If an air temperature probe is being used to influence the atmospheric correction factor, and it is being used ‘after the intercooler’ the correction factor will increase due to the hotter air temperatures.

Taking the same car and fitting a highly efficient intercooler where the intake air temperature remains constantly low.
Measuring the air temperature again ‘after the intercooler’ in the same location this will have the effect of reducing the dyno correction factor due to the cooler air temperatures measured as a result of the efficent intercooler.

The end result of testing with the air temperature probe 'influencing' the correction factor is that this could easily lead you into believing that an up-rated intercooler (that is actually more efficient) has either made no improvement to the torque and power, or worst still it could show up as having made the result become lower.

In the real world the intake air temperature value is used by some dyno operators to make power figures become inflated.
Taking a standard car and increasing the boost pressure will nearly always make the air temperature increase even if it is only slightly.
If you allow the 'hotter' air temperatures to also influence the atmospheric correction, the gain that you have made from the boost pressure will become 'amplified'. With correct tuning more boost will typically become more power but increases in the air temperature should not be taking in account and allowed to influence to the correction factor.

Above is a single example of how air temperature measurements effect power measurement and for these reasons, it is agreed by most experienced dyno operators (certainly amongst experienced Dyno Dynamics operators ) that the intake air probe must only be used to reflect the actual intake air temperature that the engine bay is taking in (which is typically within a maximum of 10-12 degrees centigrade of the ambient temperature).
This testing method results in repeatable dyno results across different atmospheric conditions (which we have proven ourselves on multiple cars) and has the effect of showing genuine gains where they are achieved.

So how Accurate is the Dyno Dynamics Shootout Mode?

JKM are in the fortunate position of having had a customer’s 500BHP 2.0 Turbocharged (Ford YB) engine tested on both an engine dynamometer (Superflow) and then again on the JKM Dyno Dynamics rolling road.


Rolling Road Testing at JKM	2.0 YB Cosworth Engine	Engine Dyno Testing (Superflow)

ITesting in accordance with all the methods outlined above, power and torque curves are shown below.


Engine Dyno : 500BHP at 6500rpm & 500Lb/Ft at 4400rpm	JKM Rolling Road: 503.6BHP at 6250rpm & 485 Lb/Ft at 5100rpm

This is a variation from engine dyno to chassis dyno of just 3.6 BHP peak and 15 Lb/Ft peak on a 500BHP & 500Lb/Ft engine! Put another way this is a variation of 0.72% on 'peak' power and 3% on the 'peak' torque level.

Having spoken with the engine builder and tuner (Not JKM), the minor difference in torque is believed to be due to the fact that the dyno 'run' started at lower rpm’s on the rolling road, this resulted in not giving the turbocharger sufficient 'inertia' to ultimately generate the same level of torque as seen on the engine dyno.
The slightly different shapes of the power/torque curves is due to the way that power measurements are made on an engine dyno that does not have the mechanical inertia of the actual rolling road drums and the car under test's transmission etc.
An engine dyno increases the RPM in operator defined steps verses an electronic controlled ramp rate acceleration method on the rolling road – also variations of the car’s true exhaust against the engine dyno exhaust will have an influence on the 'shape' of the curves.

As a very important side note these results reinforce the accuracy of the dyno dynamics ‘calculated’ flywheel method that some people like to question.

Conclusion so far......

The debate over the validity of the various Atmoshperic correction methods will always remain, but they are the only way to make realistic comparisons of your engine on different days. A dyno that is based at high altitude can not and will not produce as much power as a dyno at sea level, there is simply less oxygen for an engine to burn at high altitudes - what is less obvious are the other conditions that effect the power measured.

Do not confuse Atmospheric correction with transmission loss calculations (Power loss from flywheel to the actual tyre contact patch of the wheel) for which each rolling road manufacture evaluates and calculates themselves.
There are no Standardised ISO, DIN or equivalent corrections in use to calculate transmission losses. Some dyno manufactures evaluate this more accurately than others; again expect variations from one dyno manufacture to the next, where again operators can either introduce an error or influence the result……

As our dyno is only 2 wheel drive we can only comment on 2 wheel drive vehicles but our own estimated transmission losses for a Front wheel drive vehicle on a conventional gearbox or VAG DSG gearbox is between 15-16% and rear wheel drive cars loose between 17%-18% - again on conventional manual gearboxes.
For reasons of accuracy JKM will only ever state the power at the wheels on torque converter equipped automatic gearbox vehicles.

This report will be continued ...... coming next (More on Transmission losses)