Exagerated BHP Claims by Japanese Manufacturers

[Chui]

Interestingly enough, I saw portions of a report that indicated exaggerations in bhp claims by Toyota today. This independent company reported 272 SAE bhp for the supposed 300 bhp LS/GS430 - I think - powerplant and only 162.3 SAE bhp @ 7,250 rpm and 120.2 lb-ft torque @ 7,000 rpm for the 2ZZ-GE engine. These figures were with vehicle air induction and exhaust and "as calibrated in vehicle" spark and fueling. Only with ideal air inlet, minimum Exhaust Gas Back Pressure, Mean Best Torque spark and relative AFR of 0.9 does the car produce the numbers claimed by Toyota. Those numbers are: 178.4 bhp @ 7,500 rpm and 129.4 lb-ft torque @ 7,000 rpm. Similar results with the S2000 as well... 220 bhp as installed in the car.

To be fair, GM reported similar things when I was at GM Powertrain some time ago and it's been claimed by FoMoCo Advanced Powertrain and now by an independent source, FEV.

I was also told that this power exaggeration is common to all Japanese companies - or at least the ones who market sports cars in the US. Looking at recent past faux pas by Mazda and Nissan it makes me wonder. The report also claimed a "conspicuous loss in torque at second cam engagement" which several members' dyno sheets have shown.

Please keep in mind that these were dyno engine tests not vehicle tests on a rolling chassis dyno.

Also, the max valve lift for the Exhaust on the low cam is 7.3 mm and 9.9 mm on the high cam. On the intake side it's 7.3 mm on the low cam lobe and 11.1 mm on the high cam lobe.

The cam duration is a bit more difficult to ascertain. I'll get a scale out and interpolate.

There's 43 degrees of cam phasing on the intake - which is also listed in the SAE Paper on the 2ZZ-GE.

Valve Timing

Exhaust low lobe opens at 12 deg BBDC & closes -10 deg ATDC.

The high cam lobe opens at 38 deg BBBC & closes at 14 deg ATDC.

Low cam "advanced" opens at 11 degrees BTDC & closes at -7 deg ABDC.

Low Cam "retarded" opens at -32 deg BTDC & closes at 36 deg ABDC.

High cam lobe "advanced" opens at 39 deg BTDC & closes at 37 deg ABDC.

Hich cam lobe "retarded" opens at -4 deg BTDC & closes at 80 deg ABDC.

One other thing noted was that "torque can be improved by modifying the valve timing at switchover point."

Also, with vehicle air induction, minimum EGBP, MBT spark, AFR = 0.9 the drop in torque at switchover point is eliminated. It produced 173.4 SAE bhp and 129.4 lb-ft or torque.

All values are corrected. If I search thru the report I can find out what MBT [Mean Best Torque] is. It's the measure of the least spark advance for the most torque.

I cannot supply you with the report, but I felt some of you would appreciate the information.


Cheers.

[GTS LAID]

whats the name of it chui... i'll go buy it...

[oldster]

Wow, what a revelation, misstating HP (both ways) has been going on since the advent of the printing press........er.......was it the car? :)

[Chui]

"whats the name of it chui... i'll go buy it..."

It's a report that was generated by FEV. They will not sell it to you. The research was paid for by several manufacturers and they offered the report to them. The reports are unreal in their depth. The F20C report is 77 pages long, the 2ZZ-GE report is over 60 pages long and the BMW M5 report is a book. Really. I know they are considering testing the K20A and I wish they'd test the BMW M3 powerplant. I think that they'll test anything that Ford is willing to pay for, actually. Maybe they test other cars for other manufacturers. The more I think about it the more I feel this is how it is done.

[Griffin]

I'd be curious to see what engines they tested - and if they were brand new or not. When we dynoed the 1UZ for a project I was working on new out of the box our dyno showed 265 BHP at the flywheel (mind you this was with an austrailian ECM with no emissions and a VERY freeflowing intake and exhaust). - it was published at 300 HP in american autos and the performance it delivered once in a car seemed to indicate that it was in excess of the 265 we tested it at. So we got one that was a year old and tested it with the Australian ECM and we got 305 BHP. Take that combined with the fact that any dyono's numbers can be off by a margin and it seems pretty reasonable to me. Also - dynos are showing these celis as having like 165/175 AT THE WHEEL with only a few small intake and exhaust mods and a broken in engine - that indicates that flywheel horsepower is gonna be on the order of about 180 or so when stock.

Seems like the "no break in required" machining involves some pretty tight tolerances that needed to be loosened by wear which is not something Ford, GM, or anyone else is going to do if they are going to dyno a competitor's product now is it.

As for the drop in torque on cam shift - can you say "emissions & engine protection" ? I knew you could... this is hardly surprising and in fact it shows up on the NCF for Matrix XRS - drastic changes in cam profile would probably respond MUCH better to a little extra fueling and a bit more cam advance, but thats not gonna happen. Likewise Toyota is pretty obviously going to retard the intake cam as much as it can before they apply lift because they don't want to increase risk any kind of valve contact or anything of that nature. But mainly its probably emissions. I mean hell - thats one of the reasons they stopped selling turbocharged vehicles is because its too hard to control emissions on decelleration.

No offense dude - but if you can't show me the report and the conditions of the engine that was tested then I find it hard to lend too much real world validity to what you say. Plus lets face it - it makes SENSE for a manufacturer to dyno under perfect conditions. They ALL do it when the time comes to establish BHP and torque values, not JUST the imports. I mean - only a fool would dyno their engine on a stand with a constipated exhaust and 105 degrees F ambient and then publish that as max numbers.

Anyways all I'm saying is that your not saying anything new or shocking - they didn't LIE on the numbers - I'm sure they get every bit of what they say when they test. Remember that is MAX power and torque - they shoot for ideal condidtions because to do anythign else would be... stupid. That includes fuel too BTW - they are probably using the best damn gas they can find for this - not crappy Mobil sulfur infested pump gas. I don't think thats "cheating" at all...

Griffin

[autxr]

My car dynoed at 160 hp to the wheels (in 4th gear) with only a drop in K&N, so I find it hard to believe there is 0% drivetrain loss? This was on a DynoJet.

Scott

[t2000gts]

Originally posted by autxr
My car dynoed at 160 hp to the wheels (in 4th gear) with only a drop in K&N, so I find it hard to believe there is 0% drivetrain loss? This was on a DynoJet.

Scott

heh, i said that in the other thread. that it's impossible we're only having 0-5% drivetrain loss.

i think Griffin raised an interesting point about the break in, but they must've been dynoing them like at under 10 miles or something, because i've seen good dynoes for GTS motors with as little as 400 miles on them.

i dunno, these are engine dynoes...so maybe they don't have a fan pointed at the engine? in which case, what if all the manufacturers are deriving engine horsepower claims from wheel dynoes with some accounting done for calculated drivetrain loss?

[Chui]

No, Griffin does not. FEV is paid a couple of million dollars to test vehicles [or engines in this case] and fully follow the procedures necessitated to obtain meaningful figures. To attempt to belittle them is a bit foolhardy I think.

General Motors have test data supporting Toyota, in particular, exaggerating bhp figures. Ford Motor Company have test data supporting the same thing. In both instances the cars are broken in on a chassis roll dyno for 1,000 miles with no WOT runs. That, too, is outlined in all reports. BTW, FEV is one of the independent testing facilities for European power verifications which are far more strict than what we have here.

The losses for a transmission are variously quoted as being "5% to 10%". I dunno what they are and I'm not aware of any independent testing to verify the losses so I must state "5% to 10%".

Suffice it to say that the BMW M5 and M3 make the advertised power. As do the Porsche 996 and 996 turbo. Ditto for the V-12 M-B powerplant. Same thing for the Viper GT-S and the Corvette Z06. Engines that do not meet their advertised power: Honda and Toyota engines. Go figure.

"I mean - only a fool would dyno their engine on a stand with a constipated exhaust and 105 degrees F ambient and then publish that as max numbers."

There are specific procedures - SAE Test Procedure - as well as JIS and DIN [not to mention the new Euro test procedure] that are supposed to be strictly maintained. Each have correction factors and each are supposed to use the following:

Vehicle Induction System
Vehicle Exhaust or Vehicle Exhaust Backpressure
Vehicle Calibration


As far as whining about "not seeing the report and conditions" I've given you what I have and it really matters not what the temps were as you know that there are correction factors. If you carefully re-read what I initially posted, both the 2ZZ-GE and F20C make their advertised power only when they have:

ideal induction system - NOT the vehicle induction system
REDUCED exhaust backpressure - NOT the vehicle backpressure.
Slewed Air:Fuel Ratio - leaned out to precisely 0.9.
MODIFIED spark.


OBVIOUSLY, these are not the conditions of the engines that are in customer cars. And just so you don't think that I believe my ITR is making 195 bhp stock I do not. Never did, actually. It's officially listed as "189 bhp" in Great Britain and makes an equivalent 188 bhp in Deutschland. Now, why would the car make less power in Europe with better fuel? It wouldn't.

[oldster]

I guess I don't understand what you are trying to say. If the test you saw had the 2ZZ at 162 with a 10% drivetrain loss then that would put it at 178 so what is the problem. Am I missing something other than the fact all of your facts come from a report that noone here can collaborate?

[Chui]

Originally posted by oldster
I guess I don't understand what you are trying to say. If the test you saw had the 2ZZ at 162 with a 10% drivetrain loss then that would put it at 178 so what is the problem. Am I missing something other than the fact all of your facts come from a report that no one here can collaborate?

You may wish to re-read the conditions of the testing. Nowhere did I mention anything about a VEHICLE-level test...

Can YOU accurately verify anything about the claimed ENGINE performance of your car, oldster? Hmmm, I didn't think you could. And neither can anyone else here including myself. If you'd like a copy of the report, oldster, may I suggest you contact FEV and give them your credit card number. I'm sure they'd oblige you then YOU can give us your expert analysis since you're a powertrain engineer. We'll just sit here and wait.

:rolleyes:

[oldster]

You are right, I misread and was incorrect in my assumptions against what you had stated. It is interesting that by looking at the dynos done on member cars there is a difference of 6% between cars running stock (or no butterfly). If you take the average of those cars and assume a drivetrain loss of 10% the mean HP would be 173. How many engines did FEV evaluate to arrive at their conclusions?

As for your last paragraph I will accept that as your usual response to anyone questioning the veracity or your statements.

[GTS LAID]

I do believe there is one person that can verify ENGINE only data......

either way the situation is like this:

engine broken in for 1000 miles without WOT with:



"
the vehicle induction system
the vehicle backpressure.
Fuel Ratio - Original car ratio
Original spark.
"

and the car puts out 162.5 @ 7250

here are stock vehicles with wheel hp:

this is a 2000 2zz with 9000 miles on it:
showing approx. 163@ a tad below 7000rpm

http://www.newcelica.org/other/dyno/jesse%20il/pic1.jpg

The AEM stock dyno shows 164@7000+

and theres another dyno in the dyno list showing 11,000 miles,

with max hp averaging 163 @ around 7250...


So i think i'm missing something like oldster...

an advertised (non modified vehicle) engine with 1000 miles runs 162.5@7250 (according to your data)

stick that engine in a chassis, hook it up to the tranny, and put a dyno under the wheels... and you get 163-164 at anything from 6900-7500

it couldnt have been THAT much colder to compensate for entire powertrain loss... and even then, denser or not.. they were running precise ratios... but it looks like we're looking at somewhere in the range of 0-5% loss... i'm not a powertain engineer but i think that the flywheel/clutch and gear/output shaft combo would suck a little more than that.

[Chui]

I've seen S2K dyno plots that if one assumes a 10% driveline loss and it appears to make somewhere around the advertised power. But I also know that I can fudge the dyno numbers by strapping the vehicle tighter on the rollers... I have no idea who is taking the data nor do I know how well their equipment is calibrated.

I worked in Engine Dyno Development and, let me tell you, it can be a pain in the ass [not necessarily calibrating the dyno, but fuel flow benches must also be looked at as well as spark calibration] especially when one wishes to compare one dyno to another.

More to the real point is that the German and American engines do not show any evidence of fraudulent claims of advertised SAE bhp. Yes, the Cobra had a snafu in which the cylinder head castings were out of print, plus there was lots of flash in the intake manifold. With properly specced components - no mapping changes, no exhaust backpressure changes, no fuel changes - the engines did, in fact, produce the advertised SAE bhp. The German engines produce exactly what is claimed. To aks some of your questions [and mine, too] what bhp and torque values did the BMW M5, M3 and 740D vehicles produce vs the engines on the dyno? That would give us some idea what the driveline losses are for a FR configuration. No values exist to my knowledge like that. Perhaps, somewhere in FoMoCo I can dig up those numbers and ASSUME them to be true for any FF configuration. That's about all I can do. Speaking with persons at SVT they claim that driveline losses are on the order of 5%. Go figure.

"As for your last paragraph I will accept that as your usual response to anyone questioning the veracity or your statements."

If you'd like a copy of the report, oldster, may I suggest you contact FEV and give them your credit card number. I'm sure they'd oblige you then YOU can give us your expert analysis since you're a powertrain engineer.

Why not give 'em a phone call and ask? I'm not at liberty to share the document with you. Simple and plain. I've shared as much as I could. Period.

[oldster]

"But I also know that I can fudge the dyno numbers by strapping the vehicle tighter on the rollers... I have no idea who is taking the data nor do I know how well their equipment is calibrated. "


We could say the same thing about the data that you have presented. It would seem odd that all of the dynos that we have mentioned whether in error or not are all going the opposite direction from the data you have presented. Hmmmm......a conspiracy?

[Griffin]

Originally posted by Chui
No, Griffin does not. FEV is paid a couple of million dollars to test vehicles [or engines in this case] and fully follow the procedures necessitated to obtain meaningful figures. To attempt to belittle them is a bit foolhardy I think.


No offense but I've never worked with FEV - heckl I haven't even heard of them. I could pay some bum down on the corner to analyze my exhaust emissions by sucking on my tailpipe but that doesn't mean his results are going to mean anything. Corporations regularly waste large amounts of money on worthless subcontractors.

General Motors have test data supporting Toyota, in particular, exaggerating bhp figures.

Then why is the Pontiac VIBE advertised as having 180 HP? You telling me GM is intentionally falsifying their advertisements too? I thought the great and noble GM doesn't do that?

Ford Motor Company have test data supporting the same thing. In both instances the cars are broken in on a chassis roll dyno for 1,000 miles with no WOT runs. That, too, is outlined in all reports. BTW, FEV is one of the independent testing facilities for European power verifications which are far more strict than what we have here.

Okay - first off 1K miles doesn't mean squat in a Toyota or Honda with an alternate Cam lobe if you aren't actually driving it. If they aren't taking it to its limits Lift never has a chance to engage and you don't break in the high lift cam lobes, slipper, mechanism, valves, etc. And even if you did 1K is only 20 hours at 50 MPH. Thats not even close to breaking in. By way of example I tracked gas milage on my Tundra when I had it very closely because I was expensing all the gas and had to log it anyhow. The engine didn't really free up and start making full power and getting maximum gas mileage untill about 7K - 7.5 K. This is consistent with our findings when we dyno-ed the new and used 1UZ when we were trying to supercharge it. (privately owned engine dyno BTW not a chassis dyno) There is a BIG difference when you put some wear on the Toyota engines.

The losses for a transmission are variously quoted as being "5% to 10%". I dunno what they are and I'm not aware of any independent testing to verify the losses so I must state "5% to 10%".

average at 7.5% - thats basically 13 HP out of 180.. Hence 167... take it up towards 10% and you get 18hp loss for a number of 162.... somehow independent Chassis Dynos all over the country are consistently baselining stock GTS's at 163 / 164 HP. So is that a conspiracy by Toyota too? Maybe they paid all these Chassis dynos off? Or maybe they have created a transmission with almost no loss since you claim the Celica is 172HP and on a chassis Dyno it seems to put out over that ?

Suffice it to say that the BMW M5 and M3 make the advertised power. As do the Porsche 996 and 996 turbo. Ditto for the V-12 M-B powerplant. Same thing for the Viper GT-S and the Corvette Z06. Engines that do not meet their advertised power: Honda and Toyota engines. Go figure.

Some proof please besides for your say so in the face of solid evidence to the contrary? - Oh yeah theres this mysterious study that you can't possibly show us... But the scanned and submitted dyno sheets that HAVE been sumbitted and we CAN see are all innacurate??? I mean - I could say that every ford I've ever been in is a raving peice of crap (which is true BTW) but without proof why should I expect anyone to believe me?

There are specific procedures - SAE Test Procedure - as well as JIS and DIN [not to mention the new Euro test procedure] that are supposed to be strictly maintained. Each have correction factors and each are supposed to use the following:

Vehicle Induction System
Vehicle Exhaust or Vehicle Exhaust Backpressure
Vehicle Calibration


As far as whining about "not seeing the report and conditions" I've given you what I have and it really matters not what the temps were as you know that there are correction factors. If you carefully re-read what I initially posted, both the 2ZZ-GE and F20C make their advertised power only when they have:

ideal induction system - NOT the vehicle induction system
REDUCED exhaust backpressure - NOT the vehicle backpressure.
Slewed Air:Fuel Ratio - leaned out to precisely 0.9.
MODIFIED spark.


OBVIOUSLY, these are not the conditions of the engines that are in customer cars. And just so you don't think that I believe my ITR is making 195 bhp stock I do not. Never did, actually. It's officially listed as "189 bhp" in Great Britain and makes an equivalent 188 bhp in Deutschland. Now, why would the car make less power in Europe with better fuel? It wouldn't.

All this having been said there are also variances from engine to engine. Perhaps this company had the misfortune to buy a Celica that got a badly unbalanced low powered engine and then they didn't break it in fully before testing. Or maybe their Dyno is just in need of some recalibration. I really don't know what they did or how they did it But I'm not about to give credence to a singe study by a single company on a single vehicle, especially when I can't even see the data. If you want to post the whole study I'd love to read it. Otherwise I'm afraid I just find you claims a little hard to swallow.

Griffin

[t2000gts]

i bet toyota and honda both say that ford and GM are overrating their cars too. don't ask how, they'll find some way of proving it. and vice-versa.

[Chui]

I spoke with some individuals who also dyno engines independently who also happen to work with Mugen, the Honda tuning arm of Honda. They also confided that the ITR makes "closer to 185 bhp than 195 bhp on a test stand with vehicle induction, exhaust and calibration".

"Oh yeah theres this mysterious study that you can't possibly show us..."

It's called proprietary information. YOU do not pay $2 mill per year to them so the report does not go to you. Period. But 390 bhp for the M5 engine [five run average] and it's advertised at 394 is pretty consistent. BTW a dyno cannot accurately measure within 2% power gains/losses.

And, yes, they do have a schedule to break-in the VVT engines in which the second lobe is engaged, but no WOTs. You make a valid point about the engine further loosening up at 7.5k miles or so, but is it ethical to quote bhp figures on a 7,500 mile [or even a 15,000 mile] engine? No. And it surely isn't ethical to report gross bhp when it should be net values.

As for statistical group size is concerned there is only one "Proof Engine" that is submitted for evaluation for 'advertised power claims' anyway, so there isn't statistical representation when Honda says that the S2K makes 240 bhp or the ITR makes 195 bhp in the US.

[oldster]

Originally posted by Chui
As for statistical group size is concerned there is only one "Proof Engine" that is submitted for evaluation for 'advertised power claims' anyway, so there isn't statistical representation when Honda says that the S2K makes 240 bhp or the ITR makes 195 bhp in the US.

I was not aware that manufacturers submit engines for proof of horsepower. To whom do they submit them?

[Griffin]

Originally posted by Chui
I spoke with some individuals who also dyno engines independently who also happen to work with Mugen, the Honda tuning arm of Honda. They also confided that the ITR makes "closer to 185 bhp than 195 bhp on a test stand with vehicle induction, exhaust and calibration".

That may be true but I haven't said word one about Honda - I'm talking about the Toyota Celica GTS 2ZZ-GE

It's called proprietary information. YOU do not pay $2 mill per year to them so the report does not go to you. Period. But 390 bhp for the M5 engine [five run average] and it's advertised at 394 is pretty consistent. BTW a dyno cannot accurately measure within 2% power gains/losses.

I'm aware of that fact - which adds to the point that your study which claimed that the Toyota is 4% under stated horesepower is silly.

And, yes, they do have a schedule to break-in the VVT engines in which the second lobe is engaged, but no WOTs. You make a valid point about the engine further loosening up at 7.5k miles or so, but is it ethical to quote bhp figures on a 7,500 mile [or even a 15,000 mile] engine? No. And it surely isn't ethical to report gross bhp when it should be net values.

Maybe its not ethical to quote at 7.5k on a domestic given the poor quality of materials and lower life expectancy of teh powertrain, but given that the average life expectancy of most Toyota engines is in excess of 150K and more than 250K if well cared for I don't see a problem with it. Its only 3% of a reasonable life expectancy. However even 3K worth of real world driving conditions should yeild a good diffrence

As for statistical group size is concerned there is only one "Proof Engine" that is submitted for evaluation for 'advertised power claims' anyway, so there isn't statistical representation when Honda says that the S2K makes 240 bhp or the ITR makes 195 bhp in the US.

Right - and given any engine production like you probably have a +/- variance then you have a dyno variance as well... That alone could add up to 10% and would allow for you to gyno a 2ZZ out of the box at 260-266HP and be within specified tolerances. Likewise if I am the manufacturer I am not going to base my numbers on an engine that is at the bottom end of the spectrum... unless its a competitor's engine.

BTW since when is this about Honda? for all I know THEY might be inflating their claims - your Integra may not make as much power as it is supposed to I have no reason to doubt that the Celica can make every bit of 180HP.

Griffin

[Chui]

"BTW since when is this about Honda?"

"Similar results with the S2000 as well... 220 bhp as installed in the car." Exceprted from the initial post

READ MORE CAREFULLY.

U seem to think this is a Honda versus Toyota debate. BOTH companies are being accused of the same sh1t.

Actually, I misspoke. Three engines are handpicked for advertised power and the information is forwarded to NHTSA, etc, etc. This allows for some fudging of the numbers. Only if the engines are removed and bench tested after the MFGR'S described break-in procedures can any legal action be taken. This is what happened to Ford ['99 Cobra] and the recent Miata fiasco.

[Griffin]

Originally posted by Chui
"BTW since when is this about Honda?"

"Similar results with the S2000 as well... 220 bhp as installed in the car." Exceprted from the initial post

READ MORE CAREFULLY.

I know you originally mentioned them - and you keep bringing them up - and you might even be right about them I don't know. I just don't see any valid substantiation that Toyota has done it.

U seem to think this is a Honda versus Toyota debate. BOTH companies are being accused of the same sh1t.

No - I just dont think you can clump them together - I recognize that they are 2 very different companies with very different histories and design and operation philosophies. Well, that and I don't really care one iota about Honda. IMO the only interesting CAR they ever made was the NSX - there has been some cool technology along the way but what you confuse for Toyota VS Honda is simple indifference. I have never owned a honda and only driven a very few. They are just booring to me...

Actually, I misspoke. Three engines are handpicked for advertised power and the information is forwarded to NHTSA, etc, etc. This allows for some fudging of the numbers. Only if the engines are removed and bench tested after the MFGR'S described break-in procedures can any legal action be taken. This is what happened to Ford ['99 Cobra] and the recent Miata fiasco.

Speaking of manufacturer's break in... your company that did this test didn't follow it. The Owners manual states that for teh 1st 1000 miles you should drive between 2000 & 4000 Rpm while not maintaining a constant speed. Hence they could not possibly have followed reccomended break in AND have broken in the VVTLi system in 1000 miles. It just takes longer with Toyotas, and that could go some way towards explaining the numbers they got.

Griffin