Drive train loss riddle. (Page 2/4)
FieroWannaBe AUG 27, 10:03 AM
https://www.hotrod.com/arti...-chassis-dyno-guide/
Bring on the dyno questions, I've been away from my Day to Day job of designing these dynos for over a year and need to get back into it.
-Dave

[This message has been edited by FieroWannaBe (edited 08-27-2020).]

pmbrunelle AUG 27, 10:18 PM
Also the first run was with a Fiero transmission, and the second run was with a Getrag F23, correct?

You should probably head over to the Cavalier forums; they might know more about the F23 efficiency.

The other question that burns:
What accuracy (+/- % error) do you require for the steady-state crank hp estimation? That will influence the suggestions that are appropriate, ranging from:

Least accurate method

-Assume 10-15% loss
-Find a loss figure for the F23
-Do dyno sweeps at different acceleration rates to try to eliminate some variables. Also do steady-state measurements.
-Use an engine dyno with a dizzy/carb
-Use an engine dyno with your computer/wiring harness running the engine. If you could get spare C203/C500 connector pigtails (body side), you could make a dyno harness that would allow your engine harness to function outside the car.

Most accurate method

[This message has been edited by pmbrunelle (edited 08-27-2020).]

theogre AUG 28, 02:17 AM
All Above maybe true or not... Many are miss a Big Problem w/ dealing w/ Factory specs.
GM and other "HP specs" are often Lies.

"True" HP etc for GM et al are done on an Engine Dyno measured at the Crank Shaft, Not at the Wheels.
Even many Very High Dollar Race Teams farm out Engine dyno testing to engine makers or labs.

Ever when True numbers, you Never get Same using a wheel dyno available to the public.
Closest you get in many places are PTO Dyno's used for "Farm" tractors but even PTOs on many tractors may not be directly tried to the engine crank.

GM et al often Lies about HP and Torque.
They play w/ HP numbers even between lines of cars to make X car look better or far worse then car Y from same maker or even car X but w/ a different Trim Level. Even easier w/ computer control engines and trans because even tho X and Y get Same drivetrain, Can and Will "Detune" one of them.
Not just for Marketing either. They often Under Rate many engine like to make numbers for EPA and Insurance reason. Lower Factory HP spec often gets Lower Premiums for people buying them.

------------------
Dr. Ian Malcolm: Yeah, but your scientists were so preoccupied with whether or not they could, they didn't stop to think if they should.
(Jurassic Park)

The Ogre's Fiero Cave

Will AUG 29, 05:44 PM

quote
Originally posted by FieroWannaBe:

If the dyno is operating at a controlled accel rate, not always the case, but for a purely inertial dyno yes...
The reality is that it is generally a blend of the two.

EDIT:
Dyno Jets tend to be installed Sans eddys, so they operate on inertial load only, mustangs and superflow use eddies to add load and slow the accel rate, measuring force against a load cell, and calculating acceleration of the calibrated dyno inertia.




What I said is independent of dyno. Faster acceleration rate = greater inertial losses.

Some dynos try to maintain constant acceleration, but then introduce a LOT more variables that have to be calibrated.


quote
Originally posted by pmbrunelle:

Also the first run was with a Fiero transmission, and the second run was with a Getrag F23, correct?

You should probably head over to the Cavalier forums; they might know more about the F23 efficiency.




There's more stuff inside an F23 than a 282, but I don't think that adds up to a material difference in measured drivetrain loss. In 4th gear, each transmission has 6 loaded bearings and two helical gear power transfers, which accounts for the vast majority of the frictional losses of the transmission. Yeah, with the F23 the engine also has to accelerate the layshaft, including the 1st & 2nd layshaft gears, but those components have a small diameter, which means very low moment of inertia, which means very low inertial loss from those components.

[This message has been edited by Will (edited 08-31-2020).]

FieroWannaBe AUG 29, 09:07 PM

quote
Originally posted by Will:
Some dynos try to maintain constant acceleration, but then introduce a LOT more variables that have to be calibrated.



This can be very true, most variation in numbers from dyno to dyno will have more to do with the setup and dyno operator than it would the actual dyno. Provided everything is properly accounted for and maintained by the operator, you can expect a reasonable repeatability run per run if the acceleration rate is controlled. There is a great deal of difference in the type of hardware, software, calibration and control that is used for an OEM or R&D dyno cell vs a hobby or sportsman/repairshop dyno cell. That's somewhere in the realm $1,000,000-$2,000,000 instrumentation vs $50,000-$100,000. Just about every popular racing type engine dyno controls acceleration rate to a commanded speed. The software and control used on chassis dynos with load units does not differ very much as far as hardware and method of operation. Beyond dyno inertia calibration from the factory, the load cell needs to be calibrated regularly, and there may be a windage loss curve, depending the load units used and load cell method of construction. Usually for us hobbyist and amateur racers, they are bending beam type load cells, and load units are usually Air-cooled Eddy Current brakes). The amount of precision needed is proportional to the amount of money spent on data collection and calibration. If someone is serious about precision: 1. A chassis dyno is poor choice to achieve repeatable numbers due to variations in tire due to temperature, and wear 2. Understanding the test system should be just as important as understanding the test article, as test conditions need to controlled as tightly as the unit under test. Some software will estimate a calculated drive train loss when commanded by the operator, and it will do this by measuring the drivetrains deceleration after a run, and the rest of the math is proprietary to most manufacturers. Generally its going to be unfair to compare and engine dyno test to a chassis dyno test, unless you run the engine on the engine stand in a configuration that mirrors the installation environment closely. Most shops do not do this. OEMs are doing this more and more as SAE testing standards are becoming more refined.

If your trying to use a chassis dyno just to chase an additional 5hp, good luck. You cant control vehicle cooling very well, and you cant control the tire slip very well. And you cant expect the results from Day 5 of the 2nd week of the 3rd month of the year to directly match Day 4 of the 1st week of the 6 month of the 2nd year with 2500 miles in between. That is poor expectation management. As a disclaimer, I have been employed by a leading manufacturing of Industrial and High Performance Dynamometers.

Edit for grammer:

Usually there is an inductive pickup to measure engine speed from the ignition, drives ratios can be entered or calculated and even verified (accuracy of inputs dictate the accuracy of results). Modern software packages offer CAN and OBDII, but they can only read as fast as they are reported on the data-bus.

[This message has been edited by FieroWannaBe (edited 08-29-2020).]

pmbrunelle AUG 29, 10:09 PM

quote
Originally posted by FieroWannaBe:
Generally its going to be unfair to compare and engine dyno test to a chassis dyno test, unless you run the engine on the engine stand in a configuration that mirrors the installation environment closely. Most shops do not do this.




quote
Originally posted by FieroWannaBe:
And you cant expect the results from Day 5 of the 2nd week of the 3rd month of the year to directly match Day 4 of the 1st week of the 6 month of the 2nd year with 2500 miles in between. That is poor expectation management.



Well there's a chain of money flow, from the car owner, to the tuning shop, then to the dyno manufacturer (correct me if there are other players).

So normally, if there's customer demand for more repeatability, that demand should go up the chain first to the tuning shops, then that demand should be transferred to the dyno folks.

The local tuning shop with whom I have the coziest relationship happens to have a Dynapack (does constant RPM in addition to ramps), so that resolves the tire issue:
https://www.youtube.com/watch?v=l8n_teEuGJk

From a friendly relationship with this shop, I would probably go there for tuning regardless of which dyno they had, but the fact that it's a Dynapack pretty much cements my decision.

I suppose that some aspect of the repeatability rests on the end-customer's shoulders; he should only bring his car to be measured on days having the same temp/baro/humidity (within some tolerance). Probably depending on where you live, you should pick a weather combination that occurs frequently throughout the year, to maximize the number of suitable dyno days in a year.
FieroWannaBe AUG 29, 11:07 PM
Weather is the ever difficult variable. But that's why correction factors exist, and why your shop should have their head around how to use them, as well as the person adjusting the vehicle engine calibration. I think we would all hope the engine should perform its best during all weather conditions. Fuel density, wet bulb, dry bulb, baro. And of course the chain of money is a factor. A sensor with .5% accuracy and .5% repeatability vs a sensor with .05% accuracy and .05% repeatability have a much larger disparity in cost than they do in accuracy. There is a bottom line somewhere.
Being a Dynapak, the operators grasp on controlling the absorbers tuning parameters are just as critical as controlling the engine. Being able to sustain a stable smooth and proper response to load can be a battle in itself at times. (There is going control loop in the software). There is the benefit of inertial based test. There is no need to adjust the load based on engine output.

pmbrunelle AUG 29, 11:47 PM

quote
Originally posted by FieroWannaBe:
Being a Dynapak, the operators grasp on controlling the absorbers tuning parameters are just as critical as controlling the engine.



Is that normally tuned once and it's good for life?

OR

Since the customer's car is part of the absorber's feedback loop, that tuning must be redone for every customer car that comes in?

I expect to have a "pre-dyno meeting" where I ask all my questions about these things (so I don't fumble around inefficiently), but while you're here I might as well ask!

If there are RPM oscillations while attempting to maintain constant speed, I suppose I'll see it on the screen of my tuning laptop.
FieroWannaBe AUG 30, 12:11 AM

quote
Originally posted by pmbrunelle:
Since the customer's car is part of the absorber's feedback loop, that tuning must be redone for every customer car that comes in?



Bingo.
You probably don't need to reinvent the wheel, but every complete system has its uniqueness.
Will AUG 31, 10:26 AM

quote
Originally posted by pmbrunelle:

The local tuning shop with whom I have the coziest relationship happens to have a Dynapack (does constant RPM in addition to ramps), so that resolves the tire issue:
https://www.youtube.com/watch?v=l8n_teEuGJk




Do all hub dynos include DUAL hub units?

A single hub dyno sounds like a recipe for blown up diffs...