Comparing engines on the "desktop"...

[kris]

From another thread...

Realistically 270hp is as much as you're gonna get out of a 4.7 stroker and maintain reliability, but the 335lbft of torque that'll go with it is very handy indeed. Check out my 4.7L medium buck "recipe" on http://www.jeep4.0performance.4mg.com/stroker.html

We have two examples of "desktop dyno" results available to us for these engines right now.

Dino's have been around for a while, and recently Flash has done some numbers for a few different cams based on the same engine specs for cam comparasions.

Dino's seem to be somewhat realistic based on comparing various CR and cam choices to the Hesco engine dyno results for 9.5:1 forged/ported/polished engines. Flash on the other hand claims his numbers for a 9.5:1 engine are high, but are on average 10% lower than Dino's numbers with lower compression ratios...

Im not nit-picking on either set of numbers, Ive used Flash's to compare different cams in an identical engine, but I know that many on the web have used Dino's numbers as a guide to build their engines.

Not having one of these programs, I really dont understand how they work.
Hesco gets 260+hp and 330+lb/ft out of a 9.5:1 ported engine on a dyno. This falls right in line with Dino's 250+hp low-buck builds, although I dont consider a ported head "lowbuck" even if done in the garage. My time is worth money...

If someone did want to compare engines, how would you achieve accurate numbers ?

Either one of you want to shed some light on how these numbers are generated in these programs ?


kris.

[Mgardiner1]

I don't know what others are using, but i've played around with "desktop dyno 2003" and my numbers are ALWAYS low. Even when punching in the same value's i've seen Flash plug into his, i come up about 20% less then he does.

I'm sure that the programmer who creates the software has tweaks to his or her own algorithm. Without knowing the factors that the program takes into account (one program might show 15% response to 4+ degree cam timing, where another only shows 6%) it would be almost impossible to rely on a "desktop dyno" to give accurate number (like flash has always insisted). However, when comparing apples to apples with the same program, and just changing things around like cam specs, or CR, i think it reflects fairly well what you can expect it to do in the real world.

Just my .02

[Flash]

it would be almost impossible to rely on a "desktop dyno" to give accurate number (like flash has always insisted). However, when comparing apples to apples with the same program, and just changing things around like cam specs, or CR, i think it reflects fairly well what you can expect it to do in the real world.

If i had true air flow number it would be more acccurite.............
If i had intake air flow numbers instead of choosing from single or dual plain manifold, it would be more accurite.......................
IF i new actual air flow numbers (500 cfm,600cfm......1100cfm EXC)It could be more accurate.

But as has been said before, if you use the same program every time, you can get some pretty good comparisoms...............
Just like you vehicle to two different (real live) Dyno's..........you will get two different numbers.

Its the comparisons of parts knot the ######


Flash

[heartlandoffroad]

All dynos will read different but you would be suprised how close they will read if they are calibrated correctly and before you make your first pull on each engine. That being said any dyno/engine builder can make the numbers come out to what ever he or she wants. This is as wrong as a person can get because it makes people think they are getting more than they really are. I've seen engine dynos be off as much as 100hp just to make the customer smile and pay his bill.

[Alex22]

One of the guys I work with used to run a SuperFlow dyno at the shop he used to work at. Under the operator's bench there is a calibration dial to set the zero (just like on a multimeter) so once in a while he would mess with a customer by turning it up when they go out to lunch. It would make about 10 foot pounds more and 15 to 20 hp more, then he would piss in their cheerio's and tell them what he did and go back and reset it. All dyno's have so many correction factor inputs that you can make the little dial (hp/torque) read whatever you would like it to do. We use a DTS engine dyno where I work and I do plan on setting up a computer and running and tuning my stroker on it.
Of the many things I don't like about desktop dyno's is that they are designed by averages and trends of engines. What will help some engines may actually hurt others and when those types of errors are compounded its possible for the program to give you a digital kick in the sack or have you incorrectly jumping for joy. Do the programs take intake velocity and plenum/runner size into account as well as calculate parasitic drag and thermal efficiency into account?

~Alex

[heartlandoffroad]

DTS I love that dyno I wish I was able to work with one all the time I use a Stuska dyno I like it but I can't do part throttle runs on it. I have to go to the local Tech school to use there DTS so I can tune a standalone setup.

[Cheromaniac]

I use Dyno 2000 and I'm able to input accurate head flow data and cam specs so I think that's why my dyno numbers come close to the real thing. Nevertheless Dyno 2000 does have its limitations (intake manifold menu, exhaust menu not specific enough) so you still need to run your Jeep on a real chassis dyno.
I ran mine earlier this year. According to the dyno operator, my 200rwhp/254rwtq translate into 248hp/306lbft at the crank. Running my stroker specs in Dyno 2000 gets me 249hp/303lbft so it's very close indeed. Maybe it would have been spot on if it could model my TB spacer.

[1bolt]

Anyone played with Engine Analyzer Pro? I only gave it long enough to realize I would have had to model the Jeep I6 using a Deisel because you have to start with one of their presets and most of their preset engines are popular V8's or V6's.

I wasn't aware Hesco's numbers were for a ported and polished head. I still think 300hp is well within reach for a stroker. I didn't used to think that (for proof read the stock V8 versus Stroker comparo thread I started). But have been changing my mind more and more as I learn more.

[Cheromaniac]

I still think 300hp is well within reach for a stroker.

Not if you're limited to 5200rpm and naturally-aspirated. If you want reliable HP, 270 is about as much as you'll get and maximum displacement will be 4.7L. That extra 30hp is not as easy to find as you think and you'll probably compromise engine longevity to reach 300hp.

[1bolt]

Yeah we aren't talking about 100k mile daily drivers or budget strokers of course. We're only talking about a little over 1hp per cube, I think there's plenty more power to be found porting than anyone has yet shown. And I've not heard details or seen dyno sheets yet of a stroker that was put together using race prep and build quality.

There are guys out there building small blocks not much bigger than our I6 that put out over one and a half HP per cube, using two valves, pushrods and flat tappet cams. without reving past 6500 RPM's I don't think there's any reason that just over one HP per cube in an I6 is out of reach.

By the way no one said anything about easy. And you'll certainly compromise engine longevity for the extra power but thats a given.

[gradon]

I think that 1hp/cu. in. w/ a naturally aspirated motor is a good goal for a stroker build. I also think that 300hp is attainable, but will take skill to get there. I want to hurry up and tune mine so that I can take it to a dyno and to the strip and put some numbers up for everyone.

[Boilermaker]

Does anyone have saved engine, short block, head specs, etc. files from engine analyzer pro for the factory 4.0L engine they can share? Downloaded the trial version, looks like its pretty functional over the trial period. I've been able to locate many of the specs thru searching; but i'm sure i'm missing enough data to completely throw off the results.

Thanks in advance

[1bolt]

I wish I did, the trouble is to make a straight six you have to modify a straight six diesel engine preset... The problem with that, is that like any Dyno simulator software, they always have to make assumptions in the engine modeling to get the "baseline" un modified engine to produce an approximate number that stays semi-accurate as you make virtual mods to it.

The "assumption" on a diesel engine scares me... certainly the fact that Diesels are fuel throttled will make volumetric efficiency calculations wrong. At least...

EApro doesn't seem to have much user support for creating an engine preset, I haven't looked for 6 or 7 months but last I checked no gasoline straight six has been modeled as a preset. Once one has it's easy enough to tweak that engines build specs to have something that will look like an AMC 4.0 or stroker.

[1bolt]

You know I never noticed there was already a Toyota Supra engine modeled at the bottom of the list, the 2JZ-GTE is an inline 6, so I'm going to use it as a base to make a 4.0 model and see if we can get factory looking numbers and curves from it. Man that is a nice little inline, I was going to swap one of these into an old 81 Corolla back in my teens.

Actually it seems I've taken too much from older dyno sims, EApro doesn't appear to actually use an underlying "assumption" model, all specs seem to be there to input all the nuts and bolts specs in an engine, I wish I had dug deeper when I first got it I wouldn't have given up on it so quick...

I'm going to start a thread over in Performance Tech and see if anyone else wants to help me make an accurate set of 4.0 files. I will make different files for the major heads, be very interesting to see how much torque varies between the 7120 head and the 0331 with the smaller higher exhaust ports.

[Boilermaker]

I wish I did, the trouble is to make a straight six you have to modify a straight six diesel engine preset... The problem with that, is that like any Dyno simulator software, they always have to make assumptions in the engine modeling to get the "baseline" un modified engine to produce an approximate number that stays semi-accurate as you make virtual mods to it.

Those modeling differences probably goes a long way to explaining all the warnings i get after a run. I'm also getting some HP numbers that are pretty far off. I'll look over the example engines again, but i don't remember seeing a straight six gasoline engine in the list. I think i started with one of the amc engines and cut the cylinders to 6; i'll try starting a diesel and see if i get better results with the numbers i've been able to find. Just getting something close enough to show the relative difference when changing a part; for example the effect of a different cam on a setup, would be useful.