gtfoxy

That works out to 12.5.

Like I said before pumping losses must outpace inertia losses. So the engine was not a torque slouch, how could it with that much compression?

This is the part of the conversation I've been waiting for. Look at torque vs HP at an elevated RPM range. We all know our engines lose torque as RPM rises to a certain point. But also as an overall area under that extended operating RPM. You eventually end up at a point where the torque will be only what the inertia of the reciprocating assembly requires. It's axiomatic operation at that point & it's power is sole dependent on RPM.

Gearing becomes your next avenue once you realize the RPM range of a given gear. If you get what I am saying.

gtfoxy

I get what you're saying. Static compression is what it is but there are two important compression ratios on a 2-stroke. Low speed dynamic & high-speed dynamic. Static is just a reference for high speed.

I also understand your train of though on bore/stroke. I thought the same thing as well. That is until I started working with ideal chambers & the limitation of those necessary mechanical requirements. CR is all about ratios, obviously, but, it's all about ratios & that means things are a certain size, like spark plugs, & the space needed to light a given mass of fuel.

When you start getting into things more indepthly, with proper chamber shapes, a square bore/stroke motor is easier to play with in this regard. You reach a point where you can get beyond a certain CR.

Another thing to consider in your short stroke scenario is piston acceleration. Comparatively, within the confines of a given rod/stroke ratio range, a short stroke will have less piston acceleration, at the wrong time. What is more important here is matching that acceleration to the energy exchange rate desired by the fuel in a given chamber & CR.

Darth_V8r

iTrader: (4)

Don't get me wrong. I very much like a square engine. I think that one reason a 408 has such powerhouse potential is that it is very nearly square. But you have to design around a 4" stroke. Meaning if you don't get heads that can move a lot of air, your peak power RPM will be lower vs a 6.0.

This is where I'm really trying to understand, and I'm becoming more and more convinced the losses are in the acceleration at the top and bottom of the reciprocating parts. Here are some patterns I'm seeing in the model, and it's not ready to generalize too much, but I think it is correct for my motor.

1. Changing to a shorter stroke on the same heads, cam seems to move the peak power RPM higher (to the right)

2. Increasing stroke seems to reduce the RPM at which peak power is made (moves it to the left)

3. Increasing airflow uniformly over the entire RPM range moves the power peak to the right

4. Cutting airflow uniformly over the entire rpm range moves power peak to the left.

5. None of the above changed where peak torque is generated.

Please note I did NOT say the power output increased or dropped. Only the RPM at which peak power was made shifted.

Darth_V8r

iTrader: (4)

Another thought on the 408. Ask yourself - if you have a perfectly square 4" engine, it's a 402.

Would you increase stroke or bore to get to a 408?

gtfoxy

So this is a hypothetical?

There are a few variables here to nail down. Same chamber volume? Same piston volume? What are our rod length choices? What cam? Are you just trying to nail down engine operating dynamics as far as piston vel./ crankshaft angle, etc?

Darth_V8r

iTrader: (4)

It was more just to illustrate a point. Your goals would determine which route you took. So, if you're talking an LS block, assuming a standard deck, you would have to change rod ratios, etc, but you'd be limited by deck height, so your longer stroke will have a lower rid ratio.

Personally, I would add bore. In part because heads flow better on a wider bore, partly because it would just plain be easier to do, and partly because I would want it to rev.

But if the motor was going in a tow truck, I think I would add stroke. I wouldn't need it to rev, I would need the torque, and I would not care if peak power was at 6200 vs 6500 rpm. Always comes down to goals though right?

gtfoxy

In the grand scheme of things a .1" or so in stroke vs bore isnt going to vastly change things. Now if you were talking a 3.0" stroke vs a 4" stroke on a 4" bore motor, that is different. The deck height will predicate a given rod with a certain pin height so that is an important consideration. If it comes to stuffing a maximum stroke, and a really small rod, because you dont have the deck hieight, you can go too far with piston dwell vs acceleration.

Darth_V8r

iTrader: (4)

OK, this is a REALLY F-ed up equation, but I think I'm getting good numbers for fuel losses due to overlap. i took the curve I derived in a previous post:

------OL Time (seconds) = Overlap degrees / 6 / RPM

the point on the curve where the slope drops below 45 degrees is 0.0008222222222222222222..... seconds.

So, to get the fuel loss factor, take overlap time and divide by 0.000822 and take that ratio to the power of the square root of 2Pi (6.28) and then multiply by square root of 2 divided by 10, so this simplifies to:

------Fuel Loss Factor = 0.14 * (OL Time / 0.000822)^2.5066

Then, to calculate how much of the fuel contributes to power generation per HP output, take the Mass Fuel and multiply by 1 - loss factor.

------Net Fuel Use = Fuel Mass (g/sec) * (1 - Fuel Loss Factor) / 0.125998 (to get to imperial units) / Measured HorsePower

I sort of need to make up a term here. Something along the lines of Brake Specific Fuel Combustion (as opposed to consumption). However, the abbreviations won't be helpful. How about a BSFC efficiency?

------BSF-Combustion / BSF-Consumption ??? Another way to say it would be Net Fuel injected per HP divided by Net Fuel Properly Burned per HP

If we use this metric, at idle, only 17% of the fuel being injected is actually getting burned in time to generate power AT IDLE. This seems absurdly low at first, but it actually seems to make sense.

Once you get into the off-idle range, the losses rapidly disappear: 70% at 1200, 86% at 1600, 92% at 2000, to the point from 2800 RPM and up, BSFC efficiency is 95%+.

Think of how your cam suddenly comes into its tune at a certain RPM. Mine actually does come in right around 2800 to 3200 RPM. Also explains the fuel smell at idle that all but vanishes with a little higher RPM.

So, a quick and dirty test. If I go to 16 degrees of overlap, I show 4000 RPM being that point at which OL time drops to below 0.0008222 seconds, which would roughly predict the cam coming into its tune around 4400-4800?

OK, let the beatings begin!

gtfoxy

I would say a good thing to call is simply Overlap fuel loss in the first equation & I like Brake Specific Fuel Ignited (cant be confused). Maybe I missed it, but did you factor in your injector timing vs opening events, to any base in calculation?

Do you think all the fuel is burned BBDC in all scenarios?

Maybe I missed this too, what is your .006" Exhaust timing, in crankshaft degrees?

Darth_V8r

iTrader: (4)

BSFI. Good call.

No, I think that some fuel is still burning as the exhaust valve opens. Under low RPM, I think raw fuel is burning in the exhaust, to be honest.

I did not factor in ignition timing. I just used my engines parameters. And realizing that the EOIT is the stopping point, not starting point for injection, fuel is injecting during overlap.

When I get back to work I'll look up my 006 EVC. Or did you just mean duration? 289 duration. The cam specifics are on my work computer.

JakeFusion

iTrader: (7)

So I haven't responded here and now this has sort of morphed into whatever it is.

But here's what I'd say about your generalizations: cam timing, manifold resonance, and valvetrain instability (in a hydraulic setup vs solid roller) can change the peak on otherwise identical motors quite significantly. So that's another thing to consider.

joecar

Add to that exhaust header resonance... and how do you factor in whether the length of the headers is correct or incorrect for the RPM range.

Also, someting I meant to bring up earlier but I forgot is: how much difference do you see from non-crossover headers vs crossover headers (given that this crankshaft is of the dual plane type), and is this something you want to include...?

( And this brings up other things, like true duals vs y-pipe, and y-pipe with equal length branches vs unequal length... )

Darth_V8r

iTrader: (4)

I'm pretty sure that I have cam timing accounted for along with most airflow dynamics. I'm using VE as a direct entry, not calculated off of any assumptions. I'm also using IVC, IVO, and EVO directly in some calculations for pumping losses.

So, I'm fairly sure that when you change cam timing by say adjusting the cam gear, the VE's will change, the three valve events will change, and the calculations will shift with them.

I'm assuming a stable valvetrain, because my thought was that any instability would result in a peak reduction anyway. I'd rather understand why an engine peaks under proper conditions.

Things like this I was pretty sure would show up in the VE anyways, or was this a mistake? For example, if you change your exhaust, then go tune it, the new VE's would account for those changes? If not, it would show up in the pumping losses, and I'd have to go back and factor that in as well, but I don't have the data for much of that.

joecar

Those things probably do show up in the tuned VE table, you are correct

musthaveLSx

http://aanddtech.com/Docs/Phoenix%20workshop.pdf

You might find this of interest to you.

musthaveLSx

correct this is an interesting thread.

joecar

I was thinking about it...

if you have crossover headers (for dual plane crank), then tuned VE would be a accurate mapping; but if you had separated headers, then tuned VE would average out those mis-sequenced cylinders and so would not be an accurate mapping.

gtfoxy

Could you explain your thought process here?

gtfoxy

I want to know everything pertaining to EV timing, in crankshaft deg, along with EGT if you have it.

joecar

Crossover headers produce evenly spaced exhaust pulse nodes with symetrical low pressure antinodes in between (pulse node spacing is 180 degrees due to crossover headers alternating the bank sequence)... so the pull on any cylinder (when exhaust valve opens) is identical for all cylinders... so actual VE will be symetrical among all cylinders.

Separated headers allow the adjacently firing cylinder pairs to disrupt the 180 degree spacing (2 pairs of pulses will be 90 degrees apart, and 2 pairs will be 270 degrees apart)... so the pull on each cylinder will be different across half of the cylinders, so there will be two VE's... but since there is only a single VE table, it must average between the two VE's (i.e. it won't be correct for either VE since it is in between the two).