I think most of the bogus high numbers come from guys screwing with the weather data. Some variation is one thing, 100 hp is another. I was at an engine building shop in Charlotte a few years back, working on a project for a magazine article. Gains weren't what they had hoped for the new parts, the dyno (SF) operator did something on the console (I wasn't watching what he did), turned a knob, something, and said "here's one for the magazine". Picked up about 20 hp over the previous pull. Not the first magazine article I have seen things like done. One reason to not put too much faith in magazine articles about perf part tests. The magazine guys usually only know what they are told.
Hard to convince the young internet racers that we don't race dynos or flow benches, and that if you make big cfm or peak dyno hp numbers your god your going to be lookin' at a lot of tail lights.
numbers from chassis dyno vs. engine dyno
Moderator: Team
Bill, Superflow did this exact test in their in house dynos in the spring of 1999 for the 9th annual AETC. They took a 476" big block and put it on an engine dyno. It made 685 flywheel h.p. It was then installed in Floyd Kiel's 3000 lb. Malibu and made a chassis dyno run where it recorded 504 RWHP. This car ran 10.15 quarter mile at Brandimere's 5880' track.MadBill wrote:We need someone to test their motor on an engine dyno, not only at WOT but at 3 or 4 specific part-throttle settings as well, then repeat on a chassis dyno...
Tony, the reason the losses are way greater than 80 h.p. is because the two factors are tire to ground friction and "torque converter" losses in an automatic transmission car. The same thing can be seen in a standard transmission car that is run in anything but the 1-1 gear (4th gear in most of them). Put your Corvette in third or second and floor it and the losses will definately be higher. My son's Chevelle made 458 rwhp when the 481" in it and had stock "4026" Chevy heads, it ran low 10.50s @ 126-28, put the aluminum 335 CNC Dart heads on it, made 561 rwhp, but only picked up and 2-4 mph (10.26 @ 130) in similar conditions. The torque converter is the biggest variable in an automatic.Tony Mamo @ AFR wrote:IMO its MUCH closer to a fixed loss than a percentage.
As you get closer to the point of driveline failure you would see a greater "delta" (the fixed loss amount), but assuming you are testing in the same RPM range and applying torque loads that the particular driveline you have will reasonably put up with (every driveline will have their own fixed delta or net loss between engine dyno and chassis dyno), you will find the delta between the two figures (crank and chassis dyno #'s) is very close to a constant.
My C5 has seen three different engines with the exact same driveline....my loss is about 80 HP which is very good but inline with an efficient manual gear box and a lightweight clutch/flywheel combo. We are talking from 440 HP, to 550 HP, to 615 HP at the flywheel....all with the same 80 HP loss to the rear wheels give or take a couple depending on which chassis dyno run you wanted to count. Im builing a 700 HP larger displacement engine right now....in 6 months or less I will have another data point that likely confirms the three previous....I expect to be in the 610-620 RWHP range when I install the new motor.
Unfortunatly its not as soon as most of us would like following this thread, but even the data I have already collected clearly points to a fixed loss theory being more correct versus a percentage loss which is BS as far as Im concerned and doesnt even make sense if you have a good grasp on what is really going on.
Other sharp people in the industry share my sentiment but I have been beat up on other boards by the masses who simply cant get the percentage loss out of their mind and the threads turn into endless drivel with both sides not conceding anything (regardless of my real world findings that most people wouldnt have had the opportunity or the time/resources to duplicate).
Anyway...I was reluctant to even throw my opinion in here but it seemed enough people were questioning the percenatge situation and rightly so IMO.
Cheers,
Tony
PS....Dont forget that every car has their own particular "delta" or net loss. If I had a 50 lb clutch/flywheel combo versus a 32 lb unit my own delta would be likely 90-95 HP....not 80. If my rims and tires were heavier (or lighter) that would change it as well, but the same driveline will show very similar losses if you are testing that driveline at the same speeds of rotation. If you changed your combo completely and went from a big motor that peaked at 6500 to a very small high output engine that peaked at say 8000, that would likely increase the delta loss because you would have increased driveline RPM which would rob more power to achieve (sustain) that type of RPM given the same mass and inertia as the previous combo. I dont think it would be dramatically more, but enough to change your delta a number of some significance (maybe my 80 goes to 92 if I have to achieve peak power 1500 RPM later as an example).
-
Tony Mamo @ AFR
- Pro
- Posts: 316
- Joined: Tue Nov 07, 2006 1:24 pm
- Location:
Harold...hsutton wrote: Tony, the reason the losses are way greater than 80 h.p. is because the two factors are tire to ground friction and "torque converter" losses in an automatic transmission car. The same thing can be seen in a standard transmission car that is run in anything but the 1-1 gear (4th gear in most of them). Put your Corvette in third or second and floor it and the losses will definately be higher. My son's Chevelle made 458 rwhp when the 481" in it and had stock "4026" Chevy heads, it ran low 10.50s @ 126-28, put the aluminum 335 CNC Dart heads on it, made 561 rwhp, but only picked up and 2-4 mph (10.26 @ 130) in similar conditions. The torque converter is the biggest variable in an automatic.
I have probably spent more time on the chassis dyno than the engine dyno and I am intimately familiar with the fact automatic transmissions (especially ones with loose converters) sap a bunch more power from crank to rear wheels. Big heavy auto trans stuff (turbo 400's, 4L80's, etc.) with loose converters can see a 150 HP loss from engine to rear wheels....most typical automatic set-ups see less than that but they are all over 100 HP (certainly a huge percentage of them)....closer to 110-115 would probably be about the norm (less with a tighter converter....more with a looser converter).
Anyway....you quoted me and I wasn't really sure why because I didnt disagree with anything you mentioned.
Once again though, once you have your own particular combination's net loss (what I refer to as the "delta"), this thread is more aimed at whether you see a 1-1 relationship of any addition gains (or losses) made from there (constant or fixed loss theory), or a percentage of those gains if you happen to subscribe to that theory.
Dyno variables, cheating (cough...cough), losses thru different types of transmissions and drivelines, etc., are all interesting topics of discussion but somewhat not related to the OP's original question if I understood it correctly.
-Tony
Edit...Looks like the OP wasnt really questioning percent versus fixed loss directly but somewhere in the middle of this thread that topic kind of blew up so I guess this thread is fair game at any type of input concerning chassis dyno's and the host of variables that type of testing presents!
There's no good reason to expect good correlation from engine dyno "A" to chassis dyno "A", and definitely no reason to expect engine or chassis dynos from different shops to correlate well with eachother.
Inertia factors, weather inputs (correct or incorrect), air makeup, electrical systems, fuel systems, and on and on and on are all variables that influence testing results.
The only way to do this is by conducting a gauge R&R on separate dynos with about 15 samples.
Otherwise any resemblance of correlation is luck of the draw.
Inertia factors, weather inputs (correct or incorrect), air makeup, electrical systems, fuel systems, and on and on and on are all variables that influence testing results.
The only way to do this is by conducting a gauge R&R on separate dynos with about 15 samples.
Otherwise any resemblance of correlation is luck of the draw.
My thoughts exactly.RW TECH wrote:There's no good reason to expect good correlation from engine dyno "A" to chassis dyno "A", and definitely no reason to expect engine or chassis dynos from different shops to correlate well with eachother.
Inertia factors, weather inputs (correct or incorrect), air makeup, electrical systems, fuel systems, and on and on and on are all variables that influence testing results.
The only way to do this is by conducting a gauge R&R on separate dynos with about 15 samples.
Otherwise any resemblance of correlation is luck of the draw.
http://www.fastchip.com/
SS/JA 4156
SS/JA 4156
How can a loose converter show a bigger loss than a tight one?hsutton wrote: A loose torque converter can show more loss on a chassis dyno than one would think. A powerglide and a real loose converter can show 25% or more loss on a chassis dyno while a tighter, large converter and a T-350, might only show 19%. The quicker you try to accelerate a converter, the more power it takes.
If the loose converter accelerated the car down the drag srip faster than the tight one, how could it not accelerate the chassis dyno faster too, therefore showing an increase in HP, vs the tight one?
If you've chassed converters around, and found the fastest one at the track, even if it's a loose one, that converter should still show best on the chassis dyno. There could be some change due to a difference in acceleration rate, from track to dyno.
But, if the chassis dyno lead you to a tight converter,(showed more HP) and the drag strip lead you to a loose converter (showed lowest ET), what good would a chassis dyno be?
Randy
I'm afraid your mistaken. Looser converters, if that is what the car needs, leave harder and run quicker, but normally slip more up high. I once lost over 60 rwhp on mine when I was still driving it on the street with loose converter compared to the stock converter, but knocked 1/2 second off my 1/4 mile times. Torque numbers looked really big with the loose converter, but that was torque multiplication, and does not indicate the engine is making more torque. The tires are seeing more.
What good is it? You can find what the engine likes. If you dyno it with a stock converter first, the way I like to do mine, you know where the actual torque peak is. That is the most usefull info that will help with stall speed selection. Otherwise, your finding what it likes for spark advance and air/fuel ratio. Also you have a power comparison to previous combinations you have tried, as long as the tires, wheels, gears and converter are the same as you tested before. As I keep trying to tell people, it's tool for testing changes on that vehicel on that day.
Trying to compare chassis dyno numbers, from somebody in Florida, to somebody else's in, say Canada, is almost comical in the real world.
What good is it? You can find what the engine likes. If you dyno it with a stock converter first, the way I like to do mine, you know where the actual torque peak is. That is the most usefull info that will help with stall speed selection. Otherwise, your finding what it likes for spark advance and air/fuel ratio. Also you have a power comparison to previous combinations you have tried, as long as the tires, wheels, gears and converter are the same as you tested before. As I keep trying to tell people, it's tool for testing changes on that vehicel on that day.
Trying to compare chassis dyno numbers, from somebody in Florida, to somebody else's in, say Canada, is almost comical in the real world.
http://www.fastchip.com/
SS/JA 4156
SS/JA 4156
-
Tony Mamo @ AFR
- Pro
- Posts: 316
- Joined: Tue Nov 07, 2006 1:24 pm
- Location:
Yep...
Loose converters slip and do not produce as much power on the chassis dyno and therefore usually trap lower MPH as well (which is always the ultimate indicator of power in your timeslip). However due to the fact the engine immediatly flashes to the meat of its power curve, the car will almost always leave harder and ET better assuming it has the traction to take advantage of that. Most of your ET gains are achieved at the bottom of the track and thats where a loose converter shines.
Sometimes a slightly tighter converter will 60 foot close to the same but then run harder down the last half increasing trap speeds noticably and slightly improving the ET as well. That converter would show better on the chassis dyno if you were evaluating RWHP.
With torque converters you are alway juggling how loose versus how efficient and its tough to get both....some are better than others though....thats for sure.
-Tony
Loose converters slip and do not produce as much power on the chassis dyno and therefore usually trap lower MPH as well (which is always the ultimate indicator of power in your timeslip). However due to the fact the engine immediatly flashes to the meat of its power curve, the car will almost always leave harder and ET better assuming it has the traction to take advantage of that. Most of your ET gains are achieved at the bottom of the track and thats where a loose converter shines.
Sometimes a slightly tighter converter will 60 foot close to the same but then run harder down the last half increasing trap speeds noticably and slightly improving the ET as well. That converter would show better on the chassis dyno if you were evaluating RWHP.
With torque converters you are alway juggling how loose versus how efficient and its tough to get both....some are better than others though....thats for sure.
-Tony
I really like the lockup converters like are used on the new cars. With stock type engines and limited rpm potential the lockups do a good job of increasing big end speeds while keeping the engines in their best power producing range. Probably not the best idea for an all out 9500 rpm engine but they work real good on the near stock LS-2 Chevy engine. My son's TBSS runs 1.5 mph faster with the converter locked up.
But it would be quicker, off the bottle, if it did not have a lock up converter. They are heavier, and slow the vehicle down. I've picked up nearly a tenth going to non-lock up converters. Nitrous and turbo charged vehicles often run quicker with the converter locked. Buick GNs and the little Syclone pickups usually resond well to that. 4L60E equipped GM trucks since 1993 lock the converter at WOT at 75 mph from the factory, but they run quicker if they don't.hsutton wrote:I really like the lockup converters like are used on the new cars. With stock type engines and limited rpm potential the lockups do a good job of increasing big end speeds while keeping the engines in their best power producing range. Probably not the best idea for an all out 9500 rpm engine but they work real good on the near stock LS-2 Chevy engine. My son's TBSS runs 1.5 mph faster with the converter locked up.
http://www.fastchip.com/
SS/JA 4156
SS/JA 4156
The Trailblazer would have run it's best time if it had been able to lock the converter on the 11.362 run, with the extra torque on the juice. Not much difference though, maybe a couple of thousandths along with the 1.5 mph pick up. On a run the following weekend, with better air, it went 11.41 and he had to lift when it started spinning about 50' out. On "motor only" runs it is quicker unlocked. I think lock up was occuring somewhere north of 112 mph, or about 6000 in third gear with the nitrous. It lays over if not locked, (117 unlocked, 118.72 locked). I wonder how much more power it would show if the converter was locked? P.S..... Ed, how heavy is the converter you have in your car now? I assume it's an 8" ATI.
I thought the little 8" ones were about 20-21 lbs., dry. Neal Chance has some real light ones now in the new aluminum bolt together models. They are however going to be unbelievably expensive. I think i saw a post somewhere that said the 9" ones were under 20 lbs. and the 8" about 15 lbs. The question then would be how would heat affect them. That little 355" engine would surely see 5 fewer pounds of converter weight.