Dan Wendling

Does the engine RPM of the HP peak measured at the rear wheels on a chassis dyno differ from that measured at the crankshaft on an engine dyno?

davidfarmer

I gave an answer, and I'll give a reason, although in reality I don't know if it would be measurable. In an engine brake dyno, you have no inertial losses. Anytime you accelerate something rotationally, you have inertial losses. The greater the acceleration, the greater the losses. SO, as the torque starts to drop, the inertial losses begin to drop (which actually would mean more power relative to torque), so the peak would be higher.

Brake dynos don't measure acceleration, they measure load directly, so inertial losses are applicable. So they are more accurate, but not enough to make any difference since acceleration is what actually moves the car.

VTME13

My rational for saying no is this:

The shape of the hp curve is dictated by the cam, intake, and header design.

AU N EGL

YES ~ 15% difference from flywheel to rear wheels

Sargevette

I would also say no. Engine brake dyne put's a load on the engine that's suppose to simulate drive train and tires on the ground.

VGLNTE1

Load or not, the engine makes the power on the same level. Its all about the air flow. So intake and heads.

davidfarmer

I agree the actual power is the same, but the MEASUREMENT will give different results. Inertial losses are the reason you do chassis dyno pulls in a higher gear....the slower the acceleration, the closer you get to matching brake HP.

There will always be slight differences in acceleration (Dynojet) vs Brake/flywheel measurements.

Dan Wendling

I think the issue at hand is:

What is the shape of the driveline loss curve?

Pretty sure it is not a straight line (fixed loss)

So it either is a fixed precentage of the applied torque/HP as some people think or,

It is a complex function that includes a factor of the driveline rotational speeds and wheel speed. The resulting curve would show increasing loss in HP/Torque as the speed of the vehicle (or RPM in any given gear) increases.

My bet is the latter, and if so the net result is that the RPM of the peak HP will shift to the left (lower RPM) when measured at the rear wheels.

Shaka

I receive your stuff on Facebook, Cool!
First, you can't measure HP, it has to be calculated, so all answers to the pole questions are wrong. You don't even need a dyno if you have all engine and fuel specs corrected to a standard day. The owner of the dyno usually tests in the worst possible conditions to take his first reading. IE: A really hot, high humidity day with the fan off. Then he does his 'work' and proceeds to change the conditions for his second run. All of a sudden, there is a huge increase in HP with no work done on the engine. Watch for this trick. All runs on one job must be done in the same conditions to measure tuning improvements.
The torque can be measured directly at various places by different methods and HP can be calculated and corrected to standard day from ambient. Fortunately, a computer does all that for you.
Factors such as torque multiplication (axle tire diameter) and losses are included in the case of a chassis dyno. One dyno will be different to another. If corrections are made accurately, there will be no difference in HP calculation over an engine dyno. A 6 speed Vette would be in 4th gear as it is a direct couple save for diff ratio. Some tuners will use their own wheels for regular customers. If you get your engine back from a builder who has done a dyno run on it, do a chassis run with your own 'dyno wheels' and correct to his data sheet. I wouldnt trust anybodies numbers unless I calibrated the dyno myself. If a dyno reading is used as a base line for further tuning, that's all that is important. Remember that class action suit against Ford for their 'low' HP Cobra engines?

AU N EGL

How does the Butt dyno feel ?

Fast enough or do you need more.

davidfarmer

dan, I agree with you on mechanical/friction losses, but inertial losses are linked to accleration, so you lose more measured power when the torque is highest. So as the torque curve falls off (higher rpm), the inertial losses decrease.

Who knows, the inertial and friction losses may be a wash in 4th gear, but inertial losses are the reason we all want light wheels, lightweight brakes, and lightweight clutch/flywheel setups.

davidfarmer

Shaka, I agree with you also, but I think dynos are a good way to "measure" power, albeit it indirectly. Sure, if you could calculate every single engine point (perfect AFR, rpm, fuel usage, thermodynamic efficiency/cooling/exhaust losses) you could calculate power, but I think that is more prone to error than an acceleration dyno. All the dyno has to do is measure speed and time, and back-calculate from there. The math is easy, and you only need a few inputs.

CorvetteZ51Racer

Dan and David,
I agree that the friction loss and inertial loss curves are not linear. The intertial loss curve looks much like the power curve, as the inertial losses are greater when the engine's acceleration is greater. That's one of the benefits of an engine dyno when attempting to measure engine power output (which is different than usable power, which is the net output of the SYSTEM).

Will the power curve shift to the left a bit on a chassis dyno? Yes, but not enough to worry about - Maybe 100 rpm, not 500 rpm. That's largely due to the relative magnitudes of the losses compared to the engine power output.

CorvetteZ51Racer

So, a quick clarification on dyno types:
1) water brake dynos measure torque via a strain gauge on the output of the dyno, and deriving HP from the engineering conversion equation HP = TQ*RPM / 5252

2) eddy current dynos (that I have worked on) measure HP directly by measuring the voltage and amperage output from the alternator on the dyno output (basically it's a big generator), and from the alternator efficiency and power factor curves, calculating via HP = V*I*0.746 / (PF * Eff). Torque is back calculated from RPM and HP.

3) Chassis dynos measure the RPM rate of change of a known rotational mass (the drums in the ground), and calculate torque. If engine RPM is known via an inductive input into the dyno computer, then HP can be calculated. Even wonder why for the longest time you couldn't get HP numbers when you dyno'd a diesel, only torque? No RPM input.

Reasons for putting the car in 4th gear on a chassis dyno:
1) wheel spin is less likely. Wheel spin on a dyno is bad for data and dangerous
2) most people don't know the ratio of the non-direct gears in the transmission, so to simplify the life of the guy operating the dyno, it's put into direct drive (so only the diff ratio is needed). These are needed for taking out the driveline torque multiplication of the gear sets to determine the engine torque.


David,
I totally agree, and for those who want an engine R&D engineering lesson, keep reading. For those who want a reason to drink or need to fall asleep, keep reading.

So, if you want to calculate the engine output without any test data (in other words, theoretically), David has hit all of the high points of the data you need. However, that pesky little thermodynamic efficiency number is damned near impossible to theoretically calculate for a given engine. To understand why, keep reading (Now may be a good time to go get that glass of Scotch).

Thermodynamic efficiency is a function of the following variables:
1) Dynamic compression
2) air mass induced
3) fuel mass induced
4) air temperature
5) fuel temperature
6) homogenaeity of the AF mixture
7) heat transfer from the piston, cylinder wall and cylinder head
8) ignition location in the chamber
9) number of ignition sources in the chamber
10) ignition timing
11) flame front temperature (which is a function of a whole crap full of things)
12) flame front speed (which is a function of a whole crap full of things)
13) pumping losses on the intake
14) pumping losses on the exhaust
15) temperature gradient in the exhaust system
16) blah blah blah blah blah

Ok, so since you can't really calculate all of that crap - here's what is done in R&D for high end racing engines and lots of production engines.

There are three major areas of concern in engine analysis - cylinder pressure, friction losses, and pumping losses. To determine friction losses, you have to know the other two (unless you pull the motor apart and the bearings are in pieces. Then friction loss is calculated as "too high").

Want to measure the cylinder pressure and pumping losses? Ok, here goes:
1) need piezo-crystal pressure transducers mounted in the cylinders
2) need an optical encoder (for crankshaft position) mounted on the crank
3) if you're doing this for a 10k RPM NASCAR Cup Motor, you now need a Data Acquisition system that can measure 960,000 pressure readings per second, and properly time sync them to the appropriate crank position. Oh, and the pressure sensors have a histeresis that requires that they be re-zeroed every engine cycle, or 83 times per second - each.

From the crank position data, you can calculate the cylinder volume for each cylinder at any given time. If you take that and plot the cylinder volume vs the cylinder pressure, you now have something that looks oddly like the good old Otto Cycle from your thermo class - except it has no points or edges, and the flat line at the bottom (intake and exhaust stroke) isn't flat. Take the integral of the curve in the intake and exhaust stroke and you have the pumping losses. Take the integral of the rest of the curve and you have the cylinder pressure developed through combustion (also known as Indicated Mean Effective Pressure - IMEP). Take the sum of those two items, and subtract the measured power from the dyno, and you have the friction loss curve. Take all of this data, and build a computer simulation to model the engine's performance calibrated through the use of the measured data, and now you can start running accurate simulations of what will happen when you move the cam a few degrees.

Oh, btw, such a DAQ system costs about $500k first cost, without the dyno or the engineer needed to interpret the data.

Shaka

I think Corvette Z51 Racer got a little carried away.
Almost any dyno can tell you whether you're gaining or loosing which is all that counts. An inertial chassis dynos are proliferating because they are cheap. I would stay clear of them. It's the type of dyno PT Barnham would have if he were alive today. A sustained load chassis dyno costs 5 times more.

An acceleration or inertial chassis dyno, really ought to be named a "calculating dyno" since it attempts to measure power output by calculating power based on the amount of time required to accelerate the dyno rollers from one speed to another. This is possible when the weight of the rollers (actually the moment of inertia) is calculated into the equation, and mathematically it sounds good. The problems come with actual application and a number of inconsistent variables that cannot be built into the computation. By its very design, a power test on an acceleration dyno is a very short test, (Cheap=more profit) lasting only a matter of seconds, and during the test, the vehicle's engine and drive train are in a constant state of transition. If you choose to use this contraption, use 4th gear only. You are SOL with an auto.

The rollers should weigh about the same as the car. Aint going to happen.

The problem comes when this dynamometer is used to determine a peak power level which is what you are interested in. Torque, which is what we’re trying to measure, equals rotational inertia, which is a function of the drum that you’re are accelerating, times angular acceleration, which you calculate from how quickly the RPM is changing. The acceleration dyno has no way of accurately computing the moment of inertia of the vehicle's drive train, consequently, the peak power number generated will always be inaccurate. It will be inconsistent also so you won't be able to measure any tuning changes. Different dynos will give different results also. POS.

A sustained-load chassis dyno does not calculate the test vehicle's power output. Instead, it measures power output directly by imparting an electrical load on the rollers and measuring torque. It can sustain this load indefinitely to allow conditions to stabilize on the test vehicle. It can take readings at any desired engine speed or roller speed to exactly determine a power curve and the peak power output RPM.

You put your Vette in 4th(1:1) and go for it.. Similarly, engine RPM, wheel speed, and roller RPM can all be monitored simultaneously to immediately identify any tire slippage on the rollers. A sustained-load chassis dyno is simply more accurate. On a sustained-load chassis dyno, the weight of the rollers has no significance since the load is usually measured at a steady speed with the load imposed on the rollers. This also means the weight of your vette is insignificant.

Get actual engine dyno specs for the engine you have just installed into your Vette and calibrate the your Eddy current dyno to those specs.

Make sure you make your comparisons using the same dyno, under identical conditions, or corrected to the same standards. You also need to use the same dyno operator otherwise you are pissing into the wind.

Find a reputable operator and don't go anywhere else. Make sure his dyno meets the following criteria:

Accuracy.
Repeatability.
To tune all aspects below and above the curve we need the ability to hold at steady state under load with no wheel slip.
Worldwide Standardized factory settings (No fudging).


Don't get ripped. Well, not untill you win the race, that is.

Shaka

Mmmm.

You can do all that in one shot. Fueling.

The arguments for and against chassis dynos has always been which one is correct and which one is not. This is how you can check on your tuner dude.

If he hands you a 500 HP wheel dyno sheet you have to question the following:

What was the fuel injector duty cycle at peak power?
What was the fuel pressure?
What is the fuel injector size?
The accuracy of his equipment.

Let us start with fueling. How about fuel specs and BSFC? One pound of 93 octane unleaded gas can support a certain amount of horsepower. The amount of fuel used must jive with the HP reading. Less power can and probably will be made.

There is not an engine tuner in the world that can make more power out of that one pound of 93 octane unleaded fuel than what it can support. Watch for engine outputs higher than what the fuel can deliver. Does he have the equipment to measure brake specific fuel consumption ? BSFC is the ratio between the engine's fuel mass consumption and the crankshaft power it is producing or simply put; BSFC is the fuel used to produce the HP he claims your engine is putting out. Determining exact B.S.F.C for a specific engine is complicated and requires an engine dyno but you can get close without it. Maybe your engine came with these specs.

Based on the amount of fuel available calculations can be made to within a small percentage of how much HP a specific fuel setup can support. Based on these calculations you now can confirm if your engine tuner and his equipment is reliable by determining how much horse power your setup can make and how correct the horsepower figures are that your tuner produces.

First we want to convert fuel injector size in cc to fuel injector size in lb.

Injector size in cc/10.5 = injector size in lbs/hr

Now we want to calculate how much HP one injector can support: (Note: This calculation indicates HP on the flywheel and not to the wheels.)

Measure injector duty cycle.

Injector size in lb/hour X injector duty cycle/bsfc =HP (Note:multiply by 8)

Factors such as humidity and temperature will affect the outcome. Ask a pilot friend to correct to a standard day or repeat runs in the future in the same conditions. STD. Day: 60'f and 29.92 " HG at sea level.

More power can be made by increasing fuel injector duty cycle or installing bigger fuel injectors or by raising the fuel pressure. When raising the fuel pressure, great care must be taken to ensure that the fuel pump can supply enough fuel at the new pressure.

Raising the base fuel pressure will require a re-calculation to determine the new injector size or flow figures.

√ New fuel pressure/old fuel pressure X old fuel rate=new fuel rate.

Now calculate how much horse power a X Lb/hr fuel injector can support at Y% duty cycle.

Injector size in lb/hr X injector duty cycle/BSCF=HP

Standard practice call for 0.5 BSFC on a N/A engine.

Injector size and duty cycle plus fuel pressure will help you determine if your HP figure is within the ballpark or not. Do the calculations and if it does not match ask your engine tuner to explain.

Fueling will always help to determine where you stand.

CorvetteZ51Racer

So first of all we're talking about two separate scenarios. You're talking about taking a car to a tuner, letting them do their thing, and then back-calculating to confirm that what they're giving you makes sense. Fair enough. However, BSFC is different for every engine configuration, and cannot be assumed to always be 0.5 for a NA engine. Sorry, it doesn't work that way. Perfect example - the Daytona Prototype engines I used to design and build were 4-valve heads with a center mounted plug (Ford 4 cam). We made the same power as the LS6 and LS2 based Chevy's, but ran typically about 10% better on fuel economy. Some of that was our off-throttle programming, but a lot of that was at WOT. If you're making the same power with less fuel, you have a better BSFC.

With your method you can get in the ballpark. However my original post was intended to show that you simply cannot accurately calculate the power curve of an engine without any test data. If you don't have some baseline data (ie real world test data) from which to back calculate some of these values that should remain fairly constant from one iteration of a *given* motor to another, you just simply won't be able to accurately calculate the new output.

My other point, which you actually inadvertantly supported, was that the accuracy that you want to get out of work is heavily dependent on the money available for the testing equipment. Yes, eddy current chassis dynos are more accurate than inertial. They are also more expensive are harder to find. Likewise, the setups used in high end engine R&D are extremely expensive, not available to the public for rental, and you'll only find a few setups like them in the US.

Shaka

I didn't fully understand Dan's question in the beginning. My first thought that he wanted to contract a tuner to do some work on his engine. I didn't want to be rude and ask him what he wanted to know that for?
I answered with a broad statement that included the pitfalls that he may encounter with dyno tuners and dynos themselves. I suggested that the accuracy of actual dyno number was not important but the change after tuning was. I think that the information that he has gained to date, including your scholarly input, may cause a refinement to his query.
To answer his question, the engine HP can be calculated on a chassis dyno taking into account all the drive train losses thus far described with appropriate atmospheric corrections. The information can be used for bragging rights or a base line for tuning. Either way the accuracy of the information is academic, it won't be correct.
The thread has since expanded into a general discussion of the use of dynos and not how a dynos work. That's the nature of threads.

If conditions are kept constant, then the difference in data after tuning is useful. That is why I suggested a data sheet from the engine builder or the manufacturer as a starting point. If you don't have such info, it is fairly easy to calculate output with a certain degree of accuracy using fuel delivery specs and comparing it to the 'measured' HP. Its a start. Fuel delivery and HP changes can be correlated with even a greater degree of accuracy. All we are concerned with, is the difference between runs. Sides, there are a myriad of tuners on this forum who would gladly impart accurate fuel delivery specs. Why, they will even sell you the parts.
I used a slide rule at college with my year long V4 OHV thesis. Ed Cole used one when he designed his 265 V8. His design came out damn near perfect when the engine was finally built. My V4 was never built, thank God. The new ZL1 engine was determined to produce 650HP with fancy computers that also used fudge factors. It was reduced to 638HP in production because of emission considerations.
Dynos as shop level are pretty sophisticated these days with a huge data bases. http://www.rototest.com/rototest-dynamometer.php
Since all the engine work has been done before a dyno pull, we are not going into the engine, therefore for tuning, the focus is on fuel quality and delivery. If you want more power, you add more fuel.
If you're making the same power with less fuel as you say, then you have VE and pumping problems or maybe even engine design problems. If you say that your BSFC has improved, you don't understand what brake specific fuel consumption is or you have worded your sentence incorrectly.

Dan Wendling

I think I must not have clearly stated my original post that started this poll.

I was simply trying to get people to think about the effect of drive line losses on the measurement (calculation to be specific) of rear wheel HP.

Here are a few points to consider:

What is the slope (rate of change) of the engine dyno HP vs. RPM curve leading to the peak HP/RPM point?

What is the slope of the drive line loss HP vs. RPM curve over the same points?

If we subtract the drive line loss curve from the engine dyno curve we will get the rear wheel HP curve.

So here we get to the real question intended with the poll:

How will the rear wheel HP curve differ from the engine dyno curve?

Multiple choice answer: The shape will change such that the peak is now at:

a lower RPM or
a higher RPM or
it wont change it will stay the same

Shaka

Dan, like in the movie, "My Cousin Vinnie", Marisa Tomei said to the judge, "It's a trick Question." The dyno only measures torque and RPM at the roller drum shaft. Further information such as engine RPM and fuel flow is taken directly from the engine. The computer interpolates the two sources of information with drive train loss factors, etc, to give you approximate engine output. The only true measurements are drum shaft torque and RPM and engine rpm. If you removed the engine from the car and placed it on an engine dyno in the same room, there is just no telling which engine HP/torque curves either produces will be higher or lower. The information from the driven wheels is far more useful however.
I have to conclude that it is a trick question and can't be answered. If you measure directly at the hub like RRI does, you actually get higher torque readings than at the flywheel because of the axle ratio in the case of a Vette in 4th.. The lower the gear the higher the torque will be also. The torque is reduced by adding the wheels and measured off the rollers. The chassis dyno doesn't know that you have different diameter wheels than the last run that's why I say have you should have your own dedicated dyno wheels and tires. What info will you use? Hub torque and axle rpm or hub torque and flywheel rpm. The chassis dyno measures torque and rpm at the drum shaft yet it receives engine rpm also. Which ever rpm you use to plot your curves will give different results. I really wouldn't bother my brain with this.

www.rototest.com Here you will find powertrain graphs if you explore this website.

The best way to tune your car is at open day at your local drag strip. Bring your tools and parts and tune to your ET.