ClubWRX Forum banner

1 - 17 of 17 Posts

·
Registered
Joined
·
95 Posts
Discussion Starter #1
Ok leg humpers...have fun with that title ;).

Really, though, this is for the old school engine freaks. I'm curious about where the scooby gets it's potential for high revs, rather than being a low end torque monster.

My calculations for piston speed point to about 5800ft/sec @ 8k rpm using 3.14 stroke and 3.91 bore (2.5Litres....No calc in front of me, either....I'm doing this off the top of my head.) I thought about 6,000ft/sec was the max for forged internals? How do the rules change w/ 4 bangers and their lighter weight? None of the formuls in my books factor in rotating mass.

Edjumucate an old skool v8'd red-nekk...puleeeze???
 

·
Registered
Joined
·
1,925 Posts
The limiting factor is the stress on the rod at TDC on the exhaust stoke. Given a rod, the mass of the piston will determine the speed(and RPM) when this stress is reached.

That's how I understand it. I hope it helps.

Also, the speed in ft/s isn't necessarily the limit. It's the speed that the equipment can handle without going out of the rod's ability to accelerate that piston to a stop and back down the bore.

Edit: Sorry I don't have my books nor equations here at the office. Any websites? I haven't thought about building engines in a while.
 

·
Registered
Joined
·
95 Posts
Discussion Starter #3
<---same thing here...all my books are belong to home!

TDC at the exhaust stroke, and actually BDC at the intake stroke ar the two tensile stresses induced on the rods. TDC/exhaust is more stressful, though. At least that's how I understand it. That 6000ft limit was mentioned in one of my books and concurs with what I found on a website or two. You'd think there would be some leeway because of a 4 bangers lighter weight assembly, though. Just wanted to throw this out there and see what bites ;).
 

·
Registered
Joined
·
347 Posts
TDC/exhaust is right before the powerstroke, so I could imagine that could be the most stress the Rod experiences. I dont really understand limiting factors though so I'll be interested if anybody posts some deeper info!
 

·
Registered
Joined
·
553 Posts
you are right about the 6000 Ft/sec. mark being about the upper limit for forged internals. With cast internals you can actually get away with higher piston speeds due to less density. The only difference between 4 bangers and V8's in their ability to rev. is a lower internal co-efficient of friction and a lighter crankshaft and the lack of pushrods. The limiting factor is still your piston velocities... your bearings will just last much longer at high speeds on a 4 banger than on a v8. The real limiting factor with subies and high revs is the head/manifold design.

*edit*

I just remembered I had this bookmarked

http://www.slowgt.com/Calc2.htm#PistSpeed

some pretty awesome calculators for finding out exactly what you could get away with....
 

·
Registered
Joined
·
95 Posts
Discussion Starter #7 (Edited)
I see what you're saying about the bearings. Now, I remember reading somewhere that main and rod journal diameters play a big role in rpm capability. Basically, a larger diameter journal has lower rpm capability, but higher strength. A smaller one can spin higher, but has less strength, etc...

If I remember correctly, subaru main bearing diameter is about 2.25 inches or something close. Rod journals are closer to 2 inches, I think. Again, I don't have one in front of me to verify that.
Any input on that phenomena? I never calculated the force of something spinning in a circle....Something interesting, it looks like our engines are balanced because the vector forces balance each other out as the engine rotates. I did get THAT far in physics. (Zem...spare my mortal existence ;))

Size and spec wise, our bottom end was closest to a 340 Chrysler in specifications. The ratios involved were closest to that. I just don't know what the link was. Coupled to thost ratios are heads just built for boost, but restricted with cam size. (SOHC heads here.) Ports and valves are nicely sized, but the cam is puny. Intake valve area is 2.84 (1.42 x2). Exhaust valve area is 2.48. Look at that real hard! BIG blocks have 2.84 inch valve area with aftermarket heads! I think a certain group of engineers had boost in mind with these suckers, even if they did get put on N/A engines only. I ain't ditchin' my roller valvetrain, either.

2.5Liter--

Bore size 3.91
Crank Stroke 3.14
Rod length 5.18

Rod/stroke ratio 1.64
Piston/stroke ratio 1.24
Rod/bore ratio 1.34

...no one has the rocker arm ratios available. I'm also looking to see if STI valves and springs work on SOHC heads. Also, those rod stresses. The two worst tensile stresses occur one right after the other. TDC on the exhaust stroke, and BDC in the intake stroke. It's like a double whammy, but the first one is worse. Adding boost lessens the results of the second one, too. More food for thought.
 

·
Registered
Joined
·
553 Posts
Journal width actually has more to do with things these days than journal diameter. Back in the days of dinosaurs basically every crank out there had to deal with the same nodular iron base. Larger diameters were of course stronger however MUCH heavier (mass increases as a square of size and such) and given the relatively primative machining capabilities of the day cranks couldn't be balanced all that well. So a lighter crank ultimately allowed higher speeds because imbalances didn't matter quite as much with less overall weight.

However these days with some of the really crazy forged steel cranks and the ability to balance the things to some pretty crazy speeds you can rev a pretty large diameter crank to very high speeds. The weakness is now how wide our bearing journals are and how well supported they are. Narrower journals will be weaker but have less friction and easier lubrication, wider journals will be stronger but have more friction and a more difficult time with lubrication. If you look at guys grenading import motors the most likely things to go are the rod/main supports and the rod/main caps. Guys aren't breaking cranks so much any more but they are doing scary things to the blocks themselves.

One great thing about our boxter motors is how close together the journals are. If you look at guys who sheer off cranks it's always in the area between the journals, very rarely on the journals themselves. When you place the journals closer there is less stress that can be induced into the area in between and sheering off a crank is now harder to do. When thinking in terms of dinosaurs, subaru journals are effectively larger than they really are.

Another really great thing about our subaru motors is the effective lack of main caps. The Lower main for one cylinder bank is basically the upper main for the opposite cylinder bank. Not only is this like having hilariously huge main caps, it also distributes force more evently across the whole structure making it harder to destroy bearing caps.

In forced induction motors valve area has much less to do with things than you'd think. You have to remember that on a NA motor all the pumping is done through the bottom end and with a forced induction motor all the pumping is really done before the head. I won't go too much into head design, but keep in mind 2 things, port flow and port velocity. Remember the basics, you want to maintain as much port velocity as humanly possible without sacrificing flow because port velocity is where we get our torque. If you can stuff massive amounts of air into a head from above the valves instead of below the valves you can get huge flow rates through small valves AND keep your port velocity at really crazy levels. Really the only reason to open up the valves in a FI motor is because the ports start to restrict flow TOO much and your cylinder pressures stop rising in conjunction with manifold pressures. You really want to run the smallest valves you can without sacrificing cylinder pressure. This is where small FI motors and big NA V8's differ the most and I'll stop now because it gets kinda crazy from here in out and I could fill MANY pages with FI head design.

Oh yeah... about the valves, STi valves are sodium filled and have a much larger stem diameter. Unless you want to start doing some crazy machine work on the valve guides stick with OEM style valves.

BTW it's cool to see another engine building nerd around these parts :)
 

·
Registered
Joined
·
95 Posts
Discussion Starter #9
So it's width, not length that counts ;). I see what you mean about the bearing spacing, too. Less overall crank length makes for more torsional rigidity. It resists being twisted like a candy-cane and fracturing because of it. I know a subaru is capable of 1,000+hp. VW's did it 30 years ago running on alcohol w/ 60psi of boost. Block strength in a scoob will come from newer piston sleeve replacements, like they have for Hondas and other more "popular blocks". It's just no one has thought the design to be worth it for scoobs yet. It's only a matter of time before someone does this. Enlarging the case-bolts migh be another issue. Keeping both 1/2's of the block together screaming above 10k rpm might be a problem. I honestly think 14,000 rpm is possible out of one of these engines. New cylinder heads are only a matter of time. Dart is just now cranking out new import cylinder heads. Perhaps some form of internal or external engine bracing is in order to give the block more rigidity. Ya know how a v8 block can torque itself in 1/2 and crack in the middle of the block? I think maybe a scoob could do the same thing. Enlarging the cylinder head bolt diameter and keeping the heads sealed to the block for 40+psi pressures might be an issue, too.

Dad used to beat the **** out of me for touching his tools. So I just read obsessively, never getting his arrogant approval of anything. I got whipped every time I tweaked something and he didn't know it. Why? Cuz I love speed...**** his ****! Tire frying throttle blips were a dead giveaway, even if I did just tweak the stock parts a bit. I take everything apart now anyway, still partially playing catch-up. He never could stop me from visualizing things, and that was a key to engineering: envisioning it first. Strangely enough, his engineering work rubbed off on me anyway....and I learned far better people skills. I have more books and computer programs on this stuff now than he ever had. From that reading, I came across Subarus looking for a good car in the snow. Being from Florida, I never looked at them there. The more I read about them and what the cars could do/were made of....it just went from there :cool:. I grew up old school style, with a pen, protractor, and compass. 3d drawings are a habit. That's what I got in trouble for mostly in middle and high school. I didn't pick fights, I drew!

Most of that reading and practice at a young age lead me to think of engines as really a giant history lesson. Cars are just airplanes with wheels. Most of the technology in a car comes from the aerospace industry anyway. A Subaru having a closer connection to that made a good impression to me as a consumer. Boxers, V's, inlines, W engines, rotaries, cam intricacies, rocker ratios and pushrod lengths, "roller" anything...you name it I've got a book on it somewhere. I don't care what anyone says, these engines are ****ing built to the hilt and just keep getting better. "Coming together" in the translation of "Subaru" means a lot.

I think mechanics make good engineers, but not the other way around. Engineers think too critically rather than just turning the damn wrench and enjoying some filth ;). The latter is me. I never got away from the influence as a kid, and now it's just a hobby I can't outrun. I don't have to hide it anymore either. That's a good feeling.
 

·
Registered
Joined
·
553 Posts
It's Cool to see companies like dart getting involved in subie hardware. I think subie engines in street cars will really start getting AMAZING in the next few years. Up until us power fiend americans really woke up to the impreza platform the most powerful subies in the world were being turned out by JUN. They were pretty impressive (~750hp). However, things like the ESX car makes the stuff that JUN was turning out look completely silly... and guys like ESX have had a full decade less development time than JUN had.

I don't think we'll ever see suby motors rev much past 10k with any semblence of reliability. The problem is that our rod ratios are simply too long. You have to remember that import blocks really aren't built to deal with alot of side loading because that would add tons of extra material to the sides of the blocks which would add lots of weight and size to the engines. Dinosaur builders didn't really have to worry about this too much so it wasn't un-common to see rod ratios > 2. Unfortunately the internal dimensions/geometry of our subie blocks don't really allow for very long crank offsets/short rods so sideloading is really the biggest barrier I see to very high revving high power engines. The old engineering addage "x + y + z, you may only pick 2" comes into play. In this instance you have revs + sideloading + cylinder pressures... pick 2.

I think guys that have a foundation in the purely conceptual world of mechanical stuff before getting any experience with the practical side of things will always do the best because they can build mental ideas for how things work and then fill in the details when they start playing with the real world stuff they start to apply all the theory knowledge and fun things happen. From what I've seen all the really GREAT mechanics out there start life as engineers and just get bored with it.
 

·
Registered
Joined
·
1,545 Posts
It would seem to me that the overbore design of our engines would be happier to rev than an overstroked design. Displacements being equal, the longer stroke at the same rpms would give a lower velocity/lighter tensile load on the piston of the overbored engine since it is travelling less distance in the same amount of time. Whether an engine is a v8 or a flat or inline 4 is not as pertinent to max rpms as the overbore vs. overstroke factor I would guess. Most modern engines have good bearing design and tensile load limits are the real limits of stress.
 

·
Registered
Joined
·
839 Posts
Not what I see

The problem is that our rod ratios are simply too long. You have to remember that import blocks really aren't built to deal with alot of side loading because that would add tons of extra material to the sides of the blocks which would add lots of weight and size to the engines. Dinosaur builders didn't really have to worry about this too much so it wasn't un-common to see rod ratios > 2. Unfortunately the internal dimensions/geometry of our subie blocks don't really allow for very long crank offsets/short rods so sideloading is really the biggest barrier I see to very high revving high power engines.
That's not the way I see it. The USDM WRX has very typical rod/stroke ratios.
High side loads on the cylinder wall/piston occur with small rod/stroke ratios <1.6, many consider 1.7 - 1.75 as ideal, few engines have R/S ratios over 1.9. In the long rod engines the rod angles are much less acute creating less thrust on the cylinder wall.

The USDM WRX has a 130.43/75 R/S of 1.739

DSM 4G63 150/88 R/S 1.704

Honda B18C and SR20DE R/S 1.58

Chevy 302 5.7/3.00 R/S 1.9
Chevy 350 5.7/3.48 R/S 1.64
Chevy 400 5.565/3.75 R/S 1.484
Chevy 454 6.135/4.00 R/S 1.533

Chrysler 318/340 6.123/3.31 R/S 1.849
'58 Chrysler 392 6.951/3.90 R/S 1.782
Chrysler 426 hemi 6.871/3.75 R/S 1.832

Ford 302 5.156/3.00 R/S 1.718
Ford 429 6.605/3.59 R/S 1.839

I think you have it backwards, the small rod/stroke ratios like in the Chevy 454, 400 are the ones that produce killer side loads. Smokey Yunick was a strong advocate of large rod/stroke ratios up near 2:1 as a result of his research on a long rod Chevy Indy engine.

Our short stroke results in relatively low piston speeds which should favor rpm. Unfortunately the large bore oversquare design is more prone to detonation than an engine like the 4G63 which is quite undersquare.

http://forums.nasioc.com/forums/showthread.php?threadid=76639


Larry
 

·
Registered
Joined
·
95 Posts
Discussion Starter #13
(before you read this...where's hotrod/Larry? I thought he was an old engine geek, too?)

Harv- I agree about overbore engines vs. overstroked. Ideally you want the longest you can get out of both plus rod length....but...pick 2. We got a long stroke, larger bore, and shortest rod out of the deal.

Well..drag racing isn't realy about reliability, TOO much. Make the parts last, yes. But you still have to push, or massage limits, better than the other guy in the next lane. Porche race engines rev well past 10k, supposedly. They still get rebuilt after every race. In aerospace terms, used in certain racing circles today, hours of use is how an engine is rated. Revs are a huge part of that wear and how much engine you have left. It's like they age faster in a shorter amount of time. Reliability depends on your circumstances. Rally over undreds of miles...or run for 7 seconds at a time with a lot higher rev limits. It's exactly that xyz thing...pick two! Comparing racing goals, it's time, distance, and speed itself (more like t/d=s, if you get the drift ;)). Lower speed for a longer time and longer distance (rally, nascar). Longer distance, shorter speed, shorter time (bonneville.). Higher speed w/ lower distance, lower time (drag). The metal pulled out of the ground is still only going to do so much. The whole point of any equation is balance. Getting as close to that balance as possible is when real world stuff comes into play....and gets a whole lot more complicated. Inefficiencies result in loss of componentry...or loss of metal comparatively.

10 years ago, when imports caught on, I was wondering why no one ever stuck a 4 banger alky indy race engine into a drag style chassis. They were cranking about 1500hp with the turbo on. So what if it's in a front wheel drive car? Now look, they're running 7's in FWD 4 bangers and still cutting times down. You know how many dumb red necks whacking to nascar DON'T know that 4 bangers can run faster than a v8 at ALL??? It's not always the engine...it's the chassis and all the math in that, too. I used to think everybody knew this stuff. They don't.

I totally agree that it's the machine work and parts used that get you there power wise. Basically, a drag engine has looser tolerances than a rally or nascar engine. Parts remain relatively the same materially, but the math changes accordingly with the goal in mind. Shorter tracks in nascar will see and engine with higher redlines and shorter gears to accelerate faster. Longer tracks will have lower redlines to make the engines last longer. But, the gearing is different to reach that higher top speed that takes longer to get to with more distance to cover it in! These conditions necessitate different airflow requirements, too. So, heads get changed around in accordance to what the bottom end has to go through. An engine making 100 drag passes or so, at 7-8 seconds a piece (theoretically) only sees a bit over 10 minutes of run time....aside from dyno runs, blown anything, idle time, or other necessary on-time outside of your goal. Being at the limits of the alloys presented to us makes more sense now. Tensile strength as a limit of power, after the machine work. I guess it involves prayer at that point??? Since metal comes out of the ground, it is subject to geologic forces too. That's what goes on, in a loose sense, to the metal inside an engine. We just reinforce it, now on a molecular level and not just machining, to push limits even further. Alloys get changed, metal gets melted and almost completely frozen. They get coated with other material and then baked. Then, perhaps even more machining! Labor gets expensive after a while. We're doing more to the parts to make them tougher and last longer, and that's always cost more money....no exceptions.

About those side tolerances analog mentioned, that's why I mentioned the thicker cylinder liners. www.dartonsleeves.com Making solid deck engines out of open decked engines and making them live. That's a huge key to maintaining those revs, past good machining of what's inside or achieving good head flow. When the internal metal holding the piston sleeve is removed, you're left with the engines casing...like an egg shell with nothing in it. Ya'll know that I'm pretty sure. Replacing it with those cylinder liners, you can fill in that entire area making a true solid deck block. It's like a machined insert as opposed to filling in the block with cement. These aren't for street engines.

ungh...time for bed. I still love this stuff though, more than anything. It's not about ego or competition to me until it hits the track...THEN I'll try my damnedest to drive it and blow your doors off. If I could just drive and nothing else, it'd be perfect. Every bit of this learning has been so I could get behind the wheel of something violent. That's the point where "street-cars" just don't cut it....and my pocket book cuts it even worse. Low revs, low speed, and low octane. "Poo" I say! If it doesn't burn your eyes and get you high, what fun is it? I'm not having any "real" fun yet and it's hell trying to work to get there for some reason. Gotta get the impreza running. The blower is just in it's downtime phase.

Need sleep first....zzzzzzzzzzzzzzzzzzzzzzz
 

·
Registered
Joined
·
2,491 Posts
Re: Not what I see

hotrod said:
Honda B18C and SR20DE R/S 1.58
OK, before this thread gets moved out of MS, most of it is more than my less than Zem-size coconut can digest, but I understand that smaller stroke enables higher RPM as Honda often decreases stroke from non-VTEC to VTEC engines for higher RPM and HP. Also I see the reasoning of easier revving of less mass of fewer cyclinder motors (4 v. 8). It's often been pointed out that the Type-R engine (B18C) has a higher piston speed than even Formula 1 engines of that era (1997-2001). I don't know how fast F1 engines revved back then, but today they rev up to 16,000 RPM. It doesn't sound like an 8,400 RPM engine (B18C) would have a higher piston speed than a 16,000 RPM engine. It's rather amazing that they are able to rev a 10 cyclinder to 16,000 RPM. I'd think if they went back to 8 cyclinders, they'd be able to rev even higher. I think most widebody commercial jet engines only rev to 12,000 RPM.
 

·
Registered
Joined
·
95 Posts
Discussion Starter #15
I'm so sure about the specifics of honda internals. As far as rev capability of those engines now, it's mostly due to the poppet valve design. There's infinite control over duration, opening and closing rate, and total lift. It's like infinitely controllable VTEC, for lack of a better term. I'd love to get my hands on something like that for a Subaru. I think they're controlled by a computer, but I don't know if motec has that capability or not. It's kind of the upper end of engine research I have yet to look into. Combine that with a high revving boxer engine...and things might get really interesting.
 

·
Registered
Joined
·
128 Posts
F1 motors before being forced to the one engine per weekend rule, typically revved the 19k rpm mark. BMW was revving 19.5k. being a V10 with only 3.0 liter displacement or 300cc per cylinder, the piston speeds are far surpassing any vtec engine.
 

·
Registered
Joined
·
553 Posts
Just about to leave for CO for the weekend so I'll make this short, sorry for the lack of explanation.

Sideloading at BDC is at it's max. with a short rod ratio with little sideloading at TDC. Sideloading is at it's max at TDC with a long rod ratio and at it's minimum at BDC with a shorter ratio.
 
1 - 17 of 17 Posts
Top