Transmission Losses
Discussion
From an earlier thread.
Sweeping statement there Dom...not sure I understand - maybe you could explain why all TVRs are the same?
Hopefully this will answer all question,sorry if this bores you all The Science of Dyno Testing
TVR_owner said:
Tvr Power said:
For all you Piston Head boys & Girls add 15% to give you flywheel figures, 15% is the true calculation formula for drive train loses on all TVRs
Dom Sweeping statement there Dom...not sure I understand - maybe you could explain why all TVRs are the same?
Hopefully this will answer all question,sorry if this bores you all The Science of Dyno Testing
Tvr Power said:
For a long time now there have been heated discussions about rolling roads and their
accuracy, testing methods, why results are different on various rolling roads etc. Having
owned 2 different types of rolling road and operated several others, I am in a position to
speak from experience as well as from a scientific perspective. What I would like to do is to
explain in depth about the science behind all of this so that you can see why it's such a
heated subject!
HOW A DYNO WORKS
First of all, what is a dyno? A dyno (which is short for dynamometer) is an instrument
designed to measure the power and torque of an engine. There are two main types -Engine
Dyno, and Chassis Dyno. For the purpose of this discussion, I will only be comparing chassis
dynos as there are several types, however a brief description of an engine dyno is also given.
Engine Dyno -This is a dyno which measures the power and torque of the engine directly at
the crank or flywheel. The engine dyno is connected to the engine on a special bench and the
engine's performance can be measured directly. There are people who claim that a hub dyno
such as Rototest is the only accurate way of measuring engine power. This is not true. The
only 100% accurate way of measuring engine power is to measure the power at the engine!
Chassis Dyno - This type of dyno is often called a rolling road because the car is either driven
onto the dyno or the dyno is attached to the wheel hubs. Either way, the power and torque is
not measured directly at the engine, but instead it is measured at the wheels or the hubs. This
means that the power and torque measured will not account for any drag loss created by the
transmission, drive train, and tyres. As a result, most chassis dynos also have the capability
to measure these drag losses and add the results to the wheel power measurements in order
to obtain the engine power. However these measurements are never 100% accurate, which
will be explained later.
Types of Chassis Dyno - There are many types of chassis dynos. For the moment I will only
discuss dynos for 2WD cars. The 3 main types are twin roller dynos, single roller dynos, and
hub dynos. Twin roller dynos have two small rollers where the wheels sit between them. A
single roller dyno has one large roller where the wheels sit directly on top. A hub dyno has no
rollers at all, but instead has 4 mini dynos that connect directly to the wheel hubs.
Types of Load -To measure engine power, a load needs to be applied. This is basically a
form of mechanical resistance which makes the engine work hard to produce maximum
power and torque. You can see this effect on a turbo car very easily. If you give full throttle
with the engine in neutral, you get very little boost because there is no load on the engine. If
you give full throttle when driving in 4th gear up a hill, you get maximum boost because there
is a lot of load on the engine. A dyno therefore has to provide a load in order to work the
engine into producing maximum power. There are two main types of load system on a dyno.
These are brake loads and inertia loads. A brake load is literally that -a braking system that
applies a braking force to counter the power produced by the engine. The braking mechanism
can be either an electrical brake retarder, a water brake, a hydraulic brake, it doesn't really
matter as long as it can provide a braking force. An inertia load on the other hand is created
by getting the engine to turn a large mass. A large mass has high inertia which will resist
changes in velocity more than a small mass. By having a very heavy roller, this will resist the
force applied by the engine to accelerate it, therefore providing the load for the engine. This
principle is exactly the same principle responsible for a large heavy car needing much more
power to be launched from standstill as fast as a much lighter car with less power. An
important note here is that some dynos such as the Dynojet can function as an inertia dyno or
as a brake dyno using the optional eddy current retarder, so you can choose which method
you want for applying load.
Measuring Power and Torque - OK, so now we need to understand how the dyno measures
power and torque. Without going into any in depth physics, there are three main ways in
which to do this. For brake dynos as described above, a torque cell is used to measure the
torque applied by the load. If you know the torque force applied at the wheels and you know
the roller speed, you can work out the engine power as measured at the wheels less the drag
loss. If you also measure the rpm, simple maths in conjunction with the roller speed gives you
the engine torque as measured at the wheels less the drag losses, and is completely
independent of the gear ratio used. In fact, if you measure the engine speed and the roller
speed, you can work out the total drive train gear ratio and hence always have engine torque
as measured at the wheels. Inertia dynos measure power by measuring the acceleration of
the large inertia roller using the basic principle that force = mass x acceleration. It's a lot more
complex than that because it's a rotating mass, but you get the idea. As with the brake roller,
measuring the engine rpm and roller rpm will give you the drive train gear ratio, so you can
work out engine power and torque at the wheels less any drag loss. Hub dynos can also use
a torque cell for measurements, but it is also possible to use the hydraulic pressure as a
means of measuring the braking force. Remember that power and torque are functions of
each other and rpm, so as long as you can measure rpm and power, you will know the torque
and if you can measure the rpm and torque, you will know the power.
Measuring Drag Losses -Phew, this is the final bit of the basic workings of the dynos. It's also
where a lot of the errors start to creep in. So hang in there, because the interesting bits will
soon be revealed! Brake dynos try to measure drag losses by measuring the negative torque
during the coast down phase of the dynamic power run (explained later) with the clutch
disengaged, so it measures the drag loss from the tyres right through the drive train to the
clutch. This drag loss torque measurement is then added to the measured wheel torque
measurement so that engine power and torque at the flywheel can be estimated. An inertia
dyno on the other hand measures the rate of deceleration during this coast down phase. The
rate of deceleration of the drum on it's own is known due to the mass of the roller, so any
increase in deceleration caused by the drag loss of the tyres and the drive train can be
measured and added to the wheel power and torque measurements to estimate engine
power. Whichever method is used, it is not possible to determine engine power 100%
accurately. At full throttle, all the transmission is meshing tightly etc, so a lot of heat is
generated which is effectively power that is lost as heat. On coast down, there is no load on
the transmission therefore less power is lost. So during coast down the drag losses measured
are lower than they were when the power was being measured at the wheels during the
power run.
THE REAL WORLD ISSUES WITH DYNOS
Air Flow -First of all, let's discuss the issue of air flow. When you drive your turbo car at full
throttle on the road, you are getting a huge amount of air flowing through the intercooler,
cooling the inlet charge from the turbo. So when you are travelling at 80 mph, the intercooler
is working very efficiently. However, on the dyno, the car is stationary, so this real world air
flow must be simulated. Here is the first reason why dynos differ wildly in power readings -it's
the size of the dyno fan!!! Yes believe it or not, most UK rolling roads have dyno fans that are
far too small in terms of air flow and air speed to simulate real world driving or even get close
to it in order for the intercooler to work efficiently. What therefore happens is that the inlet
charge temperature climbs much higher on the dyno than it would on the road. Modern
engine management systems can detect this as a dangerous condition and in order to run the
engine safely under such extreme conditions, the ECU will try to compensate for this by either
reducing boost levels, increasing fuel enrichment level, or retarding ignition advance. All of
these result in one thing -less power. Therefore, it will come as no surprise that a dyno
equipped with very large fan capable of delivering a large volume of air at 100 mph will
produce better dyno results than a dyno which uses a small fan only capable of delivering a
smaller volume of air at say 30-60 mph. The important thing to remember here is that it's not
the dyno accuracy that's in question here. It's simply the ability of the dyno to provide the best
conditions for obtaining the best power from the engine. If one dyno can provide better real
world simulation and running conditions than another, the engine will produce more power.
Simple!
Tyre Drag Loss - The next important issue is the actual design of the dyno itself, or more
importantly how the car connects to it. Let's look at the twin roller setup, an example of which
is the Sun RAM 12 as used by Power Engineering and used by ourselves (our previous
dyno). As mentioned earlier, the driving wheels sit between two rollers. First of all it's worth
pointing out that this is not a real world simulation because on the road the tyres are never in
contact in two places. What therefore happens on a twin roller dyno is that you get extra drag
loss caused by the extra friction generated by the tyre being in contact with the rollers in two
places. This also creates extra heat in the tyres, and tyre deflection that the tyre was never
designed for. This obviously affects the power readings because more of the engine power is
being transferred into heat into the tyres than normally should be. On a single roller dyno,
such as the Dynojet (our new dyno), the tyres are in contact with the roller at a single point,
just like on the road. This means that you get less drag loss caused by the tyres because less
heat is transferred into to them, resulting in better wheel power measurements. This is the
closest approximation to a real road scenario and is now being used by many car
manufacturers, using very large single drum roller dynos supplied by MAHA. Remember that
tyre pressures will also affect the power readings as this will produce different drag losses.
The hub dyno connects to the hub, therefore there are no tyre drag losses to account for. This
is why hub dynos give higher wheel power readings than single or twin roller dynos.
Tyre Slip - On many twin roller dynos you can get wheel slip because the roller surface is a
smooth metal surface. This requires the tyres to be made sticky by either heating them up
through doing a few dummy power runs until traction is established, or by applying some sort
of sticky substance to the tyres so that they don't slip. The problem with doing such dummy
power runs is that the engine starts to get hot in the process, so by the time you come to
actually measure the power, you have gone beyond the optimal conditions for maximum
power, and so the power readings will be down. This effect is made even worse by the use of
an inadequate dyno fan. This is a common problem during dyno days, or rolling road shoot-
outs! On twin roller dynos with a grip surface, this is much less of an issue. On single roller
dynos like the Dynojet, the grip surface also eliminates tyre slip. With a hub dyno there is also
no tyre slip since the connection is with the wheel hub.
Climbing out of the Rollers -Ahh, this is what many people refer to as wheel spin, except that
it's not! This is only a problem on twin roller dynos. What happens is that during a power run,
the car can move forward or backwards, which in turn causes the wheels to lose contact with
one of the rollers. If you lose contact with the roller connected to the retarder, the load will be
reduce, thus the engine will produce less power. If you lose contact with the speed roller, you
will also affect the rpm reading if you are using the roller speed matched to the rev counter.
This will give a false power reading. The same problem can be caused by not strapping the
car into the rollers correctly, again something which only applies to a twin roller dyno. Finally
bad suspension geometry can also affect twin roller dyno measurements. Any dyno operator
that asks people to sit on a car to stop wheel spin is doing so in order to prevent this problem.
None of these problems apply to a single roller dyno or a hub dyno. On a single roller dyno,
the car does not have to sit directly on top of the roller, nor does it have to sit dead square in
order to achieve accurate readings. This is where a single roller dyno shares it's accuracy
with a hub dyno.
Vehicle Readiness -A common problem, especially during dyno days, is that the transmission
oil does not get a chance to warm up properly. This causes extra drag losses in the
transmission which should not be there if the car was driven for 15 minutes. Due to the time
constraints on a dyno day and not being able to spend time warming the transmission oil, the
car which goes on first will normally have the warmest transmission oil and will therefore
produce better wheel figures than if it ran last after being stood for most of the day allowing
the transmission oil to cool down.
Ambient Conditions -Sometimes even the best fan in the world cannot overcome the
scorching temperatures on a hot humid day, so inevitably a car that is dyno tested on a very
hot day will produce much less power than when run on a cold day. Again, the effects can be
seen during a dyno day where cars that run in the morning or in the evening will stand a much
better chance of producing better power figures than one running at lunch time when the
ambient temperature is at it's warmest. Humidity and air pressure also affect the power
readings. Most dynos can compensate for different ambient conditions by applying an
industry standard correction factor to the power measurements, but this does not compensate
for excessive charge temperatures causing the ECU to reduce the engine power output. If a
dyno has an automatic weather station, then the correction will always be accurate. However
if a dyno allows you to manually enter ambient conditions, false power readings can be
recorded, which more often than not will be on the high side where convenient!
Incorrect RPM measurement -Many dynos either do not have the facility to measure rpm, or
the operator simply chooses not to. With such dynos, it is possible to match the roller speed
to the rev counter and therefore you can have an estimated rpm based on the roller speed
since the gearing is constant. However this assumed that the rev counter is accurate! Also,
this method does not take into account that the tyre diameter increases during a power run
due to centrifugal force, thus affecting the gear ratio slightly and hence the final readings.
Engine RPM should therefore always be measure rather than matched by the rev counter.
Inlet Charge Temperature Corrections -Another facility offered by many dynos is the ability to
enter the charge temperature at the start of the run and then at the end of the run. The dyno
software then compensates for this and produces figures that are corrected. This is done on
turbo cars, and can result in hugely inflated power figures of 20-30 bhp. Many dyno operators
use this to overcome the problem of climbing charge temperatures during the dyno run, but
that is no substitute for providing an adequate dyno fan. This is by far the biggest "cheat"
method used in the industry for getting high power figures.
OPERATOR ISSUES
As a dyno operator myself, you may think this is an odd thing to talk about, but having seen
how many operators work and combined with my own experience, I can share with you some
of the reasons why power figures can be affected by the dyno operator.
Incorrect Coast Down - This is a common problem during testing. In order to measure drag
loss during coast down on a manual car, the operator must dip the clutch at the end of the run
to disengage the clutch from the engine. This way, the dyno can measure the power back to
the clutch or flywheel. But what some operators do is to put the car into neutral. This has the
effect of registering lower drag losses as you are only measuring losses back to neutral. This
is a good way of reducing engine power figures when convenient to do so!
Incorrect Loading -If you put too little load on a turbo engine, it won't produce maximum
power, especially when tuned. If you put too much load on, the charge temperature will get
too hot to quickly resulting in power loss towards the end of he run, especially with a small
dyno fan. An experienced operator will know what load to apply for the best case or worst
case results.
Power Run Start -On a turbocharged engine, the longer the run, the hotter the charge
temperature and therefore the lower the power at the end of the run. Therefore if the run is
made from low rpm, the chances are that the power at the end of the rev range will be lower
than if the power runs is started from higher up the rev range. This isn't a fudge. It's actually a
way of determining if heat is the main cause of power loss during a dyno run. For example, a
dyno test may be started from 2000 rpm to 7000 rpm on the first run, but 4000 to 7000 rpm on
the second run. If the power is less on the first run, the chances are that the power loss is
heat related. Again, this problem is only made worse by a small dyno fan.
CONCLUSION
So there you have it! Most of the issues have been covered here, and if I have forgotten any
then I will add them later. What you should learn from this is that it's pointless going from one
dyno to another when you are tuning a car. You should stick to one dyno in order to minimise
the variables that affect the results. All dynos will produce different results for all the reasons
mentioned above and some dynos will be more accurate than others due to design.
Remember that the airflow capability of the dyno fan is crucial, so the bigger the better! Now
that you are armed with the facts, go and get some proper dyno testing done!
accuracy, testing methods, why results are different on various rolling roads etc. Having
owned 2 different types of rolling road and operated several others, I am in a position to
speak from experience as well as from a scientific perspective. What I would like to do is to
explain in depth about the science behind all of this so that you can see why it's such a
heated subject!
HOW A DYNO WORKS
First of all, what is a dyno? A dyno (which is short for dynamometer) is an instrument
designed to measure the power and torque of an engine. There are two main types -Engine
Dyno, and Chassis Dyno. For the purpose of this discussion, I will only be comparing chassis
dynos as there are several types, however a brief description of an engine dyno is also given.
Engine Dyno -This is a dyno which measures the power and torque of the engine directly at
the crank or flywheel. The engine dyno is connected to the engine on a special bench and the
engine's performance can be measured directly. There are people who claim that a hub dyno
such as Rototest is the only accurate way of measuring engine power. This is not true. The
only 100% accurate way of measuring engine power is to measure the power at the engine!
Chassis Dyno - This type of dyno is often called a rolling road because the car is either driven
onto the dyno or the dyno is attached to the wheel hubs. Either way, the power and torque is
not measured directly at the engine, but instead it is measured at the wheels or the hubs. This
means that the power and torque measured will not account for any drag loss created by the
transmission, drive train, and tyres. As a result, most chassis dynos also have the capability
to measure these drag losses and add the results to the wheel power measurements in order
to obtain the engine power. However these measurements are never 100% accurate, which
will be explained later.
Types of Chassis Dyno - There are many types of chassis dynos. For the moment I will only
discuss dynos for 2WD cars. The 3 main types are twin roller dynos, single roller dynos, and
hub dynos. Twin roller dynos have two small rollers where the wheels sit between them. A
single roller dyno has one large roller where the wheels sit directly on top. A hub dyno has no
rollers at all, but instead has 4 mini dynos that connect directly to the wheel hubs.
Types of Load -To measure engine power, a load needs to be applied. This is basically a
form of mechanical resistance which makes the engine work hard to produce maximum
power and torque. You can see this effect on a turbo car very easily. If you give full throttle
with the engine in neutral, you get very little boost because there is no load on the engine. If
you give full throttle when driving in 4th gear up a hill, you get maximum boost because there
is a lot of load on the engine. A dyno therefore has to provide a load in order to work the
engine into producing maximum power. There are two main types of load system on a dyno.
These are brake loads and inertia loads. A brake load is literally that -a braking system that
applies a braking force to counter the power produced by the engine. The braking mechanism
can be either an electrical brake retarder, a water brake, a hydraulic brake, it doesn't really
matter as long as it can provide a braking force. An inertia load on the other hand is created
by getting the engine to turn a large mass. A large mass has high inertia which will resist
changes in velocity more than a small mass. By having a very heavy roller, this will resist the
force applied by the engine to accelerate it, therefore providing the load for the engine. This
principle is exactly the same principle responsible for a large heavy car needing much more
power to be launched from standstill as fast as a much lighter car with less power. An
important note here is that some dynos such as the Dynojet can function as an inertia dyno or
as a brake dyno using the optional eddy current retarder, so you can choose which method
you want for applying load.
Measuring Power and Torque - OK, so now we need to understand how the dyno measures
power and torque. Without going into any in depth physics, there are three main ways in
which to do this. For brake dynos as described above, a torque cell is used to measure the
torque applied by the load. If you know the torque force applied at the wheels and you know
the roller speed, you can work out the engine power as measured at the wheels less the drag
loss. If you also measure the rpm, simple maths in conjunction with the roller speed gives you
the engine torque as measured at the wheels less the drag losses, and is completely
independent of the gear ratio used. In fact, if you measure the engine speed and the roller
speed, you can work out the total drive train gear ratio and hence always have engine torque
as measured at the wheels. Inertia dynos measure power by measuring the acceleration of
the large inertia roller using the basic principle that force = mass x acceleration. It's a lot more
complex than that because it's a rotating mass, but you get the idea. As with the brake roller,
measuring the engine rpm and roller rpm will give you the drive train gear ratio, so you can
work out engine power and torque at the wheels less any drag loss. Hub dynos can also use
a torque cell for measurements, but it is also possible to use the hydraulic pressure as a
means of measuring the braking force. Remember that power and torque are functions of
each other and rpm, so as long as you can measure rpm and power, you will know the torque
and if you can measure the rpm and torque, you will know the power.
Measuring Drag Losses -Phew, this is the final bit of the basic workings of the dynos. It's also
where a lot of the errors start to creep in. So hang in there, because the interesting bits will
soon be revealed! Brake dynos try to measure drag losses by measuring the negative torque
during the coast down phase of the dynamic power run (explained later) with the clutch
disengaged, so it measures the drag loss from the tyres right through the drive train to the
clutch. This drag loss torque measurement is then added to the measured wheel torque
measurement so that engine power and torque at the flywheel can be estimated. An inertia
dyno on the other hand measures the rate of deceleration during this coast down phase. The
rate of deceleration of the drum on it's own is known due to the mass of the roller, so any
increase in deceleration caused by the drag loss of the tyres and the drive train can be
measured and added to the wheel power and torque measurements to estimate engine
power. Whichever method is used, it is not possible to determine engine power 100%
accurately. At full throttle, all the transmission is meshing tightly etc, so a lot of heat is
generated which is effectively power that is lost as heat. On coast down, there is no load on
the transmission therefore less power is lost. So during coast down the drag losses measured
are lower than they were when the power was being measured at the wheels during the
power run.
THE REAL WORLD ISSUES WITH DYNOS
Air Flow -First of all, let's discuss the issue of air flow. When you drive your turbo car at full
throttle on the road, you are getting a huge amount of air flowing through the intercooler,
cooling the inlet charge from the turbo. So when you are travelling at 80 mph, the intercooler
is working very efficiently. However, on the dyno, the car is stationary, so this real world air
flow must be simulated. Here is the first reason why dynos differ wildly in power readings -it's
the size of the dyno fan!!! Yes believe it or not, most UK rolling roads have dyno fans that are
far too small in terms of air flow and air speed to simulate real world driving or even get close
to it in order for the intercooler to work efficiently. What therefore happens is that the inlet
charge temperature climbs much higher on the dyno than it would on the road. Modern
engine management systems can detect this as a dangerous condition and in order to run the
engine safely under such extreme conditions, the ECU will try to compensate for this by either
reducing boost levels, increasing fuel enrichment level, or retarding ignition advance. All of
these result in one thing -less power. Therefore, it will come as no surprise that a dyno
equipped with very large fan capable of delivering a large volume of air at 100 mph will
produce better dyno results than a dyno which uses a small fan only capable of delivering a
smaller volume of air at say 30-60 mph. The important thing to remember here is that it's not
the dyno accuracy that's in question here. It's simply the ability of the dyno to provide the best
conditions for obtaining the best power from the engine. If one dyno can provide better real
world simulation and running conditions than another, the engine will produce more power.
Simple!
Tyre Drag Loss - The next important issue is the actual design of the dyno itself, or more
importantly how the car connects to it. Let's look at the twin roller setup, an example of which
is the Sun RAM 12 as used by Power Engineering and used by ourselves (our previous
dyno). As mentioned earlier, the driving wheels sit between two rollers. First of all it's worth
pointing out that this is not a real world simulation because on the road the tyres are never in
contact in two places. What therefore happens on a twin roller dyno is that you get extra drag
loss caused by the extra friction generated by the tyre being in contact with the rollers in two
places. This also creates extra heat in the tyres, and tyre deflection that the tyre was never
designed for. This obviously affects the power readings because more of the engine power is
being transferred into heat into the tyres than normally should be. On a single roller dyno,
such as the Dynojet (our new dyno), the tyres are in contact with the roller at a single point,
just like on the road. This means that you get less drag loss caused by the tyres because less
heat is transferred into to them, resulting in better wheel power measurements. This is the
closest approximation to a real road scenario and is now being used by many car
manufacturers, using very large single drum roller dynos supplied by MAHA. Remember that
tyre pressures will also affect the power readings as this will produce different drag losses.
The hub dyno connects to the hub, therefore there are no tyre drag losses to account for. This
is why hub dynos give higher wheel power readings than single or twin roller dynos.
Tyre Slip - On many twin roller dynos you can get wheel slip because the roller surface is a
smooth metal surface. This requires the tyres to be made sticky by either heating them up
through doing a few dummy power runs until traction is established, or by applying some sort
of sticky substance to the tyres so that they don't slip. The problem with doing such dummy
power runs is that the engine starts to get hot in the process, so by the time you come to
actually measure the power, you have gone beyond the optimal conditions for maximum
power, and so the power readings will be down. This effect is made even worse by the use of
an inadequate dyno fan. This is a common problem during dyno days, or rolling road shoot-
outs! On twin roller dynos with a grip surface, this is much less of an issue. On single roller
dynos like the Dynojet, the grip surface also eliminates tyre slip. With a hub dyno there is also
no tyre slip since the connection is with the wheel hub.
Climbing out of the Rollers -Ahh, this is what many people refer to as wheel spin, except that
it's not! This is only a problem on twin roller dynos. What happens is that during a power run,
the car can move forward or backwards, which in turn causes the wheels to lose contact with
one of the rollers. If you lose contact with the roller connected to the retarder, the load will be
reduce, thus the engine will produce less power. If you lose contact with the speed roller, you
will also affect the rpm reading if you are using the roller speed matched to the rev counter.
This will give a false power reading. The same problem can be caused by not strapping the
car into the rollers correctly, again something which only applies to a twin roller dyno. Finally
bad suspension geometry can also affect twin roller dyno measurements. Any dyno operator
that asks people to sit on a car to stop wheel spin is doing so in order to prevent this problem.
None of these problems apply to a single roller dyno or a hub dyno. On a single roller dyno,
the car does not have to sit directly on top of the roller, nor does it have to sit dead square in
order to achieve accurate readings. This is where a single roller dyno shares it's accuracy
with a hub dyno.
Vehicle Readiness -A common problem, especially during dyno days, is that the transmission
oil does not get a chance to warm up properly. This causes extra drag losses in the
transmission which should not be there if the car was driven for 15 minutes. Due to the time
constraints on a dyno day and not being able to spend time warming the transmission oil, the
car which goes on first will normally have the warmest transmission oil and will therefore
produce better wheel figures than if it ran last after being stood for most of the day allowing
the transmission oil to cool down.
Ambient Conditions -Sometimes even the best fan in the world cannot overcome the
scorching temperatures on a hot humid day, so inevitably a car that is dyno tested on a very
hot day will produce much less power than when run on a cold day. Again, the effects can be
seen during a dyno day where cars that run in the morning or in the evening will stand a much
better chance of producing better power figures than one running at lunch time when the
ambient temperature is at it's warmest. Humidity and air pressure also affect the power
readings. Most dynos can compensate for different ambient conditions by applying an
industry standard correction factor to the power measurements, but this does not compensate
for excessive charge temperatures causing the ECU to reduce the engine power output. If a
dyno has an automatic weather station, then the correction will always be accurate. However
if a dyno allows you to manually enter ambient conditions, false power readings can be
recorded, which more often than not will be on the high side where convenient!
Incorrect RPM measurement -Many dynos either do not have the facility to measure rpm, or
the operator simply chooses not to. With such dynos, it is possible to match the roller speed
to the rev counter and therefore you can have an estimated rpm based on the roller speed
since the gearing is constant. However this assumed that the rev counter is accurate! Also,
this method does not take into account that the tyre diameter increases during a power run
due to centrifugal force, thus affecting the gear ratio slightly and hence the final readings.
Engine RPM should therefore always be measure rather than matched by the rev counter.
Inlet Charge Temperature Corrections -Another facility offered by many dynos is the ability to
enter the charge temperature at the start of the run and then at the end of the run. The dyno
software then compensates for this and produces figures that are corrected. This is done on
turbo cars, and can result in hugely inflated power figures of 20-30 bhp. Many dyno operators
use this to overcome the problem of climbing charge temperatures during the dyno run, but
that is no substitute for providing an adequate dyno fan. This is by far the biggest "cheat"
method used in the industry for getting high power figures.
OPERATOR ISSUES
As a dyno operator myself, you may think this is an odd thing to talk about, but having seen
how many operators work and combined with my own experience, I can share with you some
of the reasons why power figures can be affected by the dyno operator.
Incorrect Coast Down - This is a common problem during testing. In order to measure drag
loss during coast down on a manual car, the operator must dip the clutch at the end of the run
to disengage the clutch from the engine. This way, the dyno can measure the power back to
the clutch or flywheel. But what some operators do is to put the car into neutral. This has the
effect of registering lower drag losses as you are only measuring losses back to neutral. This
is a good way of reducing engine power figures when convenient to do so!
Incorrect Loading -If you put too little load on a turbo engine, it won't produce maximum
power, especially when tuned. If you put too much load on, the charge temperature will get
too hot to quickly resulting in power loss towards the end of he run, especially with a small
dyno fan. An experienced operator will know what load to apply for the best case or worst
case results.
Power Run Start -On a turbocharged engine, the longer the run, the hotter the charge
temperature and therefore the lower the power at the end of the run. Therefore if the run is
made from low rpm, the chances are that the power at the end of the rev range will be lower
than if the power runs is started from higher up the rev range. This isn't a fudge. It's actually a
way of determining if heat is the main cause of power loss during a dyno run. For example, a
dyno test may be started from 2000 rpm to 7000 rpm on the first run, but 4000 to 7000 rpm on
the second run. If the power is less on the first run, the chances are that the power loss is
heat related. Again, this problem is only made worse by a small dyno fan.
CONCLUSION
So there you have it! Most of the issues have been covered here, and if I have forgotten any
then I will add them later. What you should learn from this is that it's pointless going from one
dyno to another when you are tuning a car. You should stick to one dyno in order to minimise
the variables that affect the results. All dynos will produce different results for all the reasons
mentioned above and some dynos will be more accurate than others due to design.
Remember that the airflow capability of the dyno fan is crucial, so the bigger the better! Now
that you are armed with the facts, go and get some proper dyno testing done!
spitfire4v8 said:
lol great answer dom but to a question that wasnt asked. now can you please explain why do all tvrs need a 15% correction?
steve-V8s said:
Thanks for the replies.
EzyGriff
If it was 60bhp at 6000, what was that as a percentage of the engine power? if it was 15 % as suggested that would give you 400 BHP at 6000 RPM, does that sound right ?
At a guess I'd say EG's loss was estimated at more like 21 or 22%. (I don't believe the 15% constant figure.)EzyGriff
If it was 60bhp at 6000, what was that as a percentage of the engine power? if it was 15 % as suggested that would give you 400 BHP at 6000 RPM, does that sound right ?
David Baker from Puma Race Engines wrote an interesting article about it, and I'd take his approach:
http://www.pumaracing.co.uk/coastdwn.htm
http://www.pumaracing.co.uk/coastdwn.htm
Losses aren't a constant ratio across the transmission working range. I have no problem taking the peak RWBHP and applying /85*100 to get max engine power.
I would also happily accept that Dom has compared enough of our cars on engine & chassis dynos to satisfy himself that it is accurate enough?
I would also happily accept that Dom has compared enough of our cars on engine & chassis dynos to satisfy himself that it is accurate enough?
TVR Power use 15% drive line power consumption so as no to over inflated their RR performance data. Actual Power being derived at the rear wheels, the resulting power is then increased by a multiple of 15% to achieve a minimum at the flywheel figure for reference. This is why Flywheel figures vary so much. If an engine produces 250hp at the wheels, this would be quoted as 287.5hp with drive line power consumption of 37.5hp. If working in reverse 37.5hp as a percentage of assumed flywheel power would only be 13%. This is obviously too low with real drive line losses being substantially more, probrably 60hp (depending on all the variables) this equating to about 22% of flywheel figures. Interestingly I believe the component within the driveline with most loss is the tyre, therefore maintaining correct pressure is very important if wanting max power.
If you want to measure comparative performance the only way to get close is by measuring rear wheel figures on the dyno or using a datalogger to establish performance in real time in real conditions, that must be the most accurate measurement.
If you want to measure comparative performance the only way to get close is by measuring rear wheel figures on the dyno or using a datalogger to establish performance in real time in real conditions, that must be the most accurate measurement.
Over the last few years I have had the opportunity to run many engines on the engine dyno, and then on the rollers afterwards. One of my test cars has a 4.8 Litre Lucas-Sagem GEMS engine (2WD Sierra, T56 Gearbox), so it is ultra-consistent.
This motor has been on John Eales and V8D's engine dynos before it was put in the car, and on the rollers (in fourth gear) at Power Engineering (as was), Austec, Mech Repairs, TVR Power, and Shropshire Auto Services. The difference between the engine dyno and the rollers is always 21-23%, except at Mech repairs which reads 21% less than all the others.
My understanding is that TVR Power use 15% fixed, so that they are guaranteed a low figure. In this way they are never caught up in the daft power punch-ups that we occasionally see here. Very smart if you ask me.
This motor has been on John Eales and V8D's engine dynos before it was put in the car, and on the rollers (in fourth gear) at Power Engineering (as was), Austec, Mech Repairs, TVR Power, and Shropshire Auto Services. The difference between the engine dyno and the rollers is always 21-23%, except at Mech repairs which reads 21% less than all the others.
My understanding is that TVR Power use 15% fixed, so that they are guaranteed a low figure. In this way they are never caught up in the daft power punch-ups that we occasionally see here. Very smart if you ask me.
Gassing Station | Griffith | Top of Page | What's New | My Stuff