September 28, 2011

Sven Andersson has found an interesting, knowledgeable article about dynamometer types and what differentiates them. The article is written by Sam Borgman, Torque Development International, UK, and is shared with permission, unedited.

Dyno systems - A frank overview of the field

As a car enthusiast you will undoubtedly hear the term "dyno" thrown around a great deal, with all kinds of people talking about power results measured on a dyno, engines that have been tuned on a dyno perhaps fault diagnosis which has been carried out on a dyno or even after market performance parts claiming to offer dyno proven power gains. The term Dyno is short for the word Dynamometer and a dynamometer is briefly defined as a machine which is used to measure torque and rotational speed, from which power can be calculated.

Despite all of the publicity exposure and the apparent importance of the dynamometer system to the car enthusiast Iíve found that during my time in the industry very few car enthusiasts Iíve met would say that they truly understand the whole subject of dynoís and testing, and perhaps more worryingly they are completely unaware that there are good types of dyno systems which really are worthwhile seeking out and that there are types of dyno system which should probably be avoided or at least used as last choice.

In the world of Dynamometers there are two major types, the engine dyno and the chassis dyno. The engine dyno isessentially a test bed on which to run an engine away from a car chassis and at the same time take measurements of the output. These systems are primarily used by tier-1 vehicle manufacturers or research companies working specifically on engine development. They offer superior access to the engine for easy swapping of parts and they are directly coupled to the engine so measurement accuracy is potentially good, but they have downsides too. In order to use an engine dyno system the engine must first be removed from the cars chassis so for use by us in the aftermarket the engine dyno can prove undesirable for these main reasons:

  • High labour cost involved in removing and re-installing a carís engine

  • Testing of the engine only, meaning that the performance of the actual power train as a whole remains an unknown.

  • The engine will ultimately be tested in a working environment which can often be very different to that of its eventual workplace inside the cars chassis and this can give rise to significant disparities between the engine dyno results and the engines real world performance.

To better focus the scope of this article Iím going to concentrate on the type of dyno that you are much more likely to encounter, the chassis dyno, so from this point on when I use the term dyno Iíll be referring to a machine that is designed to test a complete engine, drive line and chassis combination. Iíll explain what ideally we need from a good dyno system, Iíll also try to give you a description of the common main types of dyno systems explaining briefly how they work and most importantly how the design of a system might affect your choice as to whether to use one type or perhaps another.

What do we need from a good dyno system?

  • It must have a sturdy and reliable physical interface with the chassis in order for the car chassis to transmit torque to the machine accurately and repeatable.

  • It must have a direct method of torque measurement, the fewer links in the measurement chain the better.

  • The Torque measurement system must be able to be re-calibrated and tested with dead weights if necessary to ensure measurement reliability.

  • The dyno must be able to control the rotational speed on the chassis output extremely quickly and accurately in order to control the engine speed accurately during a test to ensure test repeatability.

  • It must have accurate rotational speed measurement.

  • In order for the results to be interpreted meaningfully the data gathering hardware and software on the dyno should be capable of monitoring real time atmospheric data, for instance both the ambient and engines intake air temperatures, the ambient relative humidity and the atmospheric pressure of the test environment.

  • A good dyno should be able to record and log other data streams taken from the car at the same time as the torque output and rotational speed information in order to make it useful as a tuning and diagnostic tool.

Inertia chassis Dyno's

These machines normally either consist of a large diameter single roller which represents a very large mass or two smaller rollers which are directly connected to a flywheel. The way that these systems work is that a car chassis strapped into position on to the machine so the tyre contact patches of the drive wheels sitting either between the twin rollers or are carefully balanced on top of the single large roller. The car is then driven by a dyno operator as normal up through the gear box until the gear selected for the test is reached, the dyno operator will then use a very small throttle opening (1 or 2%) in order to lower the engine speed to the pre-desired point before fully opening the throttle in one smooth move. The cars driveline is now supplying maximum effort to the drive wheels and is trying to accelerate the large mass of the machine as quickly as possible, during this period the inertia chassis dyno only monitors the rotational speed of the rollers, no direct or in-direct torque measurement is being made at all. The test ends when the engine speed reaches either the engines rev-limiter or a pre-desired level spotted by the dyno operator who will then de-clutch the transmission at the same time as releasing the throttle letting the engine return to idle and letting the dyno slowly spin down to a stop with the transmission in neutral.

Itís then up to the software in the inertia dynoís data logging software to work out how long it actually took for the car to accelerate the known mass of the rollers and eventually the software will calculate torque figures based on that data alone. Once torque has been calculated and logged the logging software has the calculated torque numbers stored along with the measured rotational speed numbers and so can use the information to calculate a power number.

The single roller inertia dyno is quite common in the aftermarket tuning industry because the cost of the system is extremely low when compared to the average. Unfortunately this low cost of the system does as you might expect come as a result with some compromises.

The physical coupling between the chassis and the dyno is a rolling road style so it relies completely on a rolling pneumatic tyre, this provides us with some immediate problems;

  • The amount of the cars chassis output which is being eaten up by the tyres changes greatly all through the run and along with many variables such as, the inflation pressure, suspension geometry (toe, camber etc), the contact patch pressure (weight transfer during the run and tension on the straps), the actual tyre design, the rubber compound of the tyre, the tyreís temperature and the actual amount of torque being transmitted. From the measurement accuracy stand point the rolling tyre represents a swirling storm of uncontrollable and unquantifiable variables and is very far from ideal.

  • Itís a well documented fact that the contact patch of a pneumatic tyre "squirms" in operation and an element of rubber to surface slippage is always present, the amount of slip on this small scale is unfortunately an unknown from moment to moment and seriously affects the reliability of the end results.

  • Another major hurdle is getting traction to the rollers, it often becomes especially difficult when testing high power vehicles or track based competition cars which tend to run extremely high levels of wheel camber.

  • Probably the most important point to make about tyres and dynoís is that of safety, a road tyre can be seriously affected by the intense stresses involved in transmitting maximum torque to a dyno roller whilst clamped in place with high tension ratchet straps, and there is simply no way that anyone can say with any level certainty that any particular tyre which has been used to carry out a dyno test is definitely safe for road use afterwards and that in my eyes becomes a serious problem, which is all too often glossed over by the manufacturers of rolling road dynoís.

The special problems caused by including the tyres put aside for a moment, the biggest disadvantage of the inertia based systems is that by design they cannot control the rotational speed of the vehicle output, the machine can only offer a fixed resistance. This means that a chassis with higher power output will ultimately be subjected to a much shorter test with a much sharper ramp rate than the identical chassis which is producing lower power.

Why does that matter? Well for two huge reasons, firstly the driveline has a mass i.e. itís made up of heavy parts (crankshaft, flywheel, gears, drive shafts, brake discís etc) and the amount of the vehicles output which is absorbed by accelerating these components is a direct function of the driveline inertia along combined with the acceleration rate. So by not being able to control the acceleration rate it becomes pretty impossible to meaningfully compare a lower power test run against a higher power test run (i.e. low and high boost) and therefore to gain any reliable "back to back" test data. When testing turbo charged vehicles the inconsistencies in ramp rate can take an even greater toll as a very fast ramp rate can leave a high power turbo car not able to make its maximum boost levels at itís real world rpm's purely from lack of time or in some cases with larger turbo not making its true "full boost" on the dyno at all which can have disastrous consequences if the dyno is being used for tuning.

Eddy current brake chassis dyno's

These systems again either use two relatively small diameter rollers or a larger single hollow roller (which offer very little inertia) in order to transmit the output from the vehicle to the machine, the rollers themselves are normally directly coupled to an Eddy current retarder, which is essentially a large electrical disc brake which relies on electromagnetic forces rather than friction to create a braking resistance.

These dyno's are far more flexible than the straight un-braked inertia dyno's because the Eddy current retarders can be computer controlled in order to give accurate and reasonably fast reacting rotational speed control by varying the resistance of the rollers. If understood and used correctly by a dyno operator this closed loop speed control feature removes the inherent inaccuracies caused by variable ramp rates you get during testing with inertia dynoís. Good chassis speed control also allows the dyno operator enough control over the engine speed to be able to slowly and carefully navigate all of the possible speed and load conditions an engine might see in normal operation so that then a mapper can accurately calibrate an engine management system fitted to the vehicle.

These dynoís normally take a direct measurement of torque directly from the electrical retarder units which tend to literally twist on to a torque transducer (or strain gauge) as the retarder produces a resistance to control the vehicle output speed, this transducer can depending on brand of dyno be checked and re-calibrated when necessary to ensure accuracy and to correct them back in line if need be, itís certainly not uncommon at all to see torque transducers drift over time making the whole dyno system inaccurate if they are not regularly checked and trimmed in. Whilst a big step forward from an inertia dyno these twin roller braked dynoís unfortunately still rely entirely on a rolling tyre in order to transmit the chassis output to the machine so they are still subject to all of the inherent pneumatic tyre inaccuracies mentioned previously. Whatís more with the types which employ two smaller diameter rollers the stress on the tyre is elevated even further, some of the more popular brands even advocate a cantilever strapping system to be used to secure the chassis in order to gain "good traction" with the rollers in high output testing scenarios, unfortunately this method of strapping increases the pressure being applied to the tyre and the pressure becomes a function of the chassis power output and Iím sure you can imagine that in a very high power chassis testing session this will result in enormous stresses being applied to the vehicles tyres, sometimes for long periods of time, I would consider then driving on those very same tyres a serious liability after that.

The Hub chassis dynamometer

These dynoís are an evolution of the engine test bed dyno and many tier-1 companies invest in hub mounted chassis dyno systems in addition to their normal fleets of engine dynoís for the purposes of testing their completed chassisís. These dyno systems consists of separate dyno units, one to replace every drive wheel on the chassis, the dynamometer units bolt to the cars wheel studís literally taking the place of the cars wheels and once bolted up the dyno supports the cars weight entirely.

Once a vehicle has been bolted in to a chassis dyno the coupling between the vehicle and the machine is 100%, no slip or other rolling losses can occur which might skew the results, itís this elimination of the tyre from the measurement chain which gives the hub dyno the major advantages over the "rolling road" types of dyno system, some of the immediate advantages are;

  • Safety. With the car bolted solidly to the dyno it can not go anywhere, driving a high powered car through a test using a rolling road style dyno can be a hair raising experience, not ideal when your pride and joy is at stake.

  • Safety. The extremely low inertia of the rotating parts of the dyno mean that in an emergency situation the vehicle and the dyno can be immediately brought to a full stop without fear of chassis de-stabilising or leaping off of the dyno.

  • Safety. Because the wheels and tyres are removed from the chassis for testing they are not put through any stress or heat so after testing they can be re-fitted to the car and you can drive on them with confidence.

  • Measurement Accuracy. Removing the pneumatic tyre means that the chassis is able to exert force directly on to the dyno with no rubber in the equation to generate variables, basically if the chassis makes X amount of Torque at itís hubs then X amount of Torque is transferred to and measured by the dyno every time without fail, regardless of wheel camber, caster, toe settings etc.

  • Power handling. Once a chassis is bolted in to a hub style dyno the machine is then capable of measuring its entire load capacity without any additional drama. This unfortunately is not often true when using rollers as getting enough physical traction between the tyres and the rollers can be very hard or even impossible, often ending up with terrific clamping forces being put on to tyres in an attempt to make them grip.

These systems are normally much more expensive than even the most illustrious rolling road style systems and they tend to take more time to set-up on a car by car basis, but in return they offer the very best dyno testing solution for a chassis, both in terms of operator and client safety and in accuracy and reliability of the data that they produce.

The importance of a good test cell

Testing a vehicle at maximum effort whilst itís sitting stationary presents some interesting challenges in terms of providing the car with a simulated air flow ideally the same as would be expected out on the road or track. Unfortunately properly ventilated dyno cells are considered by some to be an expensive luxury and very many tuners neglect this aspect entirely, choosing to set the dyno up in the middle of their workshop or worse tucked into a closed off corner with no external ventilation and only an internal fan for air movement. The ideal test cell would be fresh air ventilated, both climate and pressure controlled and at tier-1 levels some test cells are built with that level of complexity, but in the aftermarket the running cost makes climate controlled test cells unrealistic. The best that can realistically be attained at a bearable hourly rate is very good fresh air ventilation, ideally delivering enough speed and air volume to replace all of the air in the cell every 5-6seconds or so and enough separate exhaust extraction to remove twice as much poisonous gas as is really necessary.

Atmospheric Correction

When operating in a non-climate controlled test cell itís crucially important that we have an accurate method of correcting our measured "day to day" results back to a standard day in terms of atmospheric variables. The variables that need to be measured are Ambient temperature, Intake temperature, Ambient humidity and Ambient barometric pressure, as long as these variables are measured accurately and logged and then a reliable and proven correction algorithm is used to process the numbers then itís possible to test a vehicle in the depths of winter on a low pressure foggy day then again in the height of a heat wave in the summer with high pressure and hot dry air and the results will be comparable.

The importance of the dyno operators

Lastly and most importantly you must consider the quality of the dyno operator themselves. Many aspects about dyno testing require the person responsible for operating the dyno to truly understand how the system works. They must also know how to control crucial variables in order to maintain measurement accuracy and equality in the case of back-to-back testing. The dyno operator is an extremely important variable, even a fantastic piece of dyno equipment in the wrong hands can give inaccurate results. The reverse is also true, even a fantastic dyno operator in control of a flawed dyno system can only result in flawed measurements.