1b rock

looking to build a single turbo set up for a 91 350 tpi car
stock
what size would be a good turbo for street strip application and possibly could grow some with thanks

Kingtal0n

I vote Borg Warner S364 with a large turbine option.

~750hp capable. Small enough to spool in a range happy for the TPI. 7psi of boost runs it right through the center of its map and will tolerate 20~psi enough to blow that motor into pieces.
Anything larger and it will just be laggier and make less torque down low.

Worth mentioning?: going through all the trouble to turbo the TPI- why not swap over to a 5.3 first (since you need to fab everything anyways, plus they come with good base manifolds for turbocharging fab) which is just as cheap of an engine and provides far more upgradability, power potential, and so forth. jus sayin'

Orr89RocZ

iTrader: (20)

Kingtal0n

I know the s366 is extremely popular but I can't see how its any better than the S364 given that the S364 provides more power than a typical Stock Bottom End can handle already.

For some reason in the V8 crowd everyone tends to oversize their turbochargers. It just makes the combo more laggy and nobody ever realizes it because nobody ever downgrades and compares (its really rare)

We should do some matchbot configurations for fun.

Orr89RocZ

iTrader: (20)

Kingtal0n

^^
-This is all true and great observations-

Just a couple finite physical attributes for 'tech' discussion I would bring up for fun
'too much cylinder pressure at low rpms' theory- I see this get thrown around alot and while it is good general practice to 'avoid high boost at low rpms' it is typically tuning error that causes bent rods and issues. Imagine instead of a turbo we have a blower which can produce max boost at low rpm. Also imagine the same engine being held at a steady state rpm- unchanging or even decelerating rpm on an engine dyno at max boost and low rpm, like a boat engine. Any issue caused will be related to the timing or fuel quality; it isn't the boost at low rpm that in and of itself is causing distress to the engine, 20psi is nothing compared to the pressure generated by an improperly timed spark or poor fuel quality. And if the amount of torque exceeds the capacity of the rotating assembly, then that same amount of torque would ruin the engine at ANY rpm, not just low rpm.

To put it another way.
When an engine is at 2000rpm and 20psi of boost for example, imagine if we apply negative timing, say -5* btdc. Surely this will limit torque production as the piston is already sweeping away from any climbing cylinder pressure. This is the sort of tuning theory that is often overlooked because most 'tuners' tend to start at timing numbers relevant to higher rpms and stay in that range (say from 9* to 12*) because they are not used to tuning for torque at low rpms on the same engine all the time. It is merely an matter of experience and/or proper diagnostic equipment.

One more puzzle piece:
The rate of change of the engine is far more important than the rpm when we consider the building of cylinder pressure and how it related to torque. If the rate of change is low or zero (steady state rpm) then the cylinder pressure needs to build more slowly than when the engine is accelerating quickly. We can infer from this that numerically lower gears (1st gear for example) will be more forgiving to large btdc timing values than high gears (overdrive) because the overdrive gear tends to hold dRPM closer to zero, which means the combustion gasses have more time to expand against a piston which is accelerating slowly (less volume is being 'dealt' to the expanding gasses over time) giving it more chances to "blow something apart" (a head gasket hopefully).

summary:
-If an engine has a known related torque capacity, you tune under that capacity, no matter what the rpm is, because torque at any rpm is considered as an instantaneous condition.
-when we look at engine torque, we do not concern ourselves with the RPM (Frequency) of the engine nearly as much as the rate of change (dRPM) of the engine.

Orr89RocZ

iTrader: (20)

Kingtal0n

for fun brainstorming

down and dirty rod bend/low rpm torque theory analysis apprentice approach
an engine can have torque without any RPM.

Kingtal0n

The top post his max length (15,000 characters) so I am forced to update thread with new details in a separate posts
Note that I am not very good at 'this', that is putting what is normally math formulas into regular words. The goal of these posts is to use normal words we all understand to explain higher phenomena which I have come to realize are differential equations that even textbooks about higher mechanics of materials will avoid using due to complexity.
I do not pretend to be especially great at explaining these things (part of the reason for length is I keep trying to explain the same thing a new way) however I have seen very little to none of the topics I wish to discuss anywhere on the internet in obvious places (connecting rod forces especially at low rpm due to reaction rate and other factors such as rate of change of the engine).

---- section 2 de-myths
Internal combustion engine model pdf I found if you get a chance:
https://cefrc.princeton.edu/sites/ce...ton-CEFRC1.pdf
Although it does not cover the details we are interested in atm (rod stress and high torque scenarios) it still has a LOT of information that I am sure we could also be more familiar with.

brief review of torque equation we used in section 1:
some engine radius are large, and some are small. The radius is part of an equation (the torque equation) which started off as straight line work assuming movement 12" (one foot) in some forward distance, and now with a radius is saying 6.28 feet per revolution because a rotating object 1 foot radius moves 2 pie radians in one revolution (circumference), and we set radius to 1 with 1 foot distance from the center of rotation. So if you are looking at the output of an engine with an exceptionally small or large radius different much from r=1 foot I'm trying to figure out how that would impact the results of this simple equation. In terms of rod forces, at first glance it means to me that rods on crankshafts which are small radius with equal torque to larger engine crankshafts, will experience more stress because less radius means more force is required on the end to give the same torque result at the same rpm. That checked out logically with whats obvious about torque at a distance and it explains why manufacturers use larger stroke engines to handle larger torque/towing applications.

In a way its like saying, more credit to putting the same torque output with a smaller engine. But more stress also. So it epitomizes engineering advancement in a way, if similar results can be attained using engines with smaller radius reliably doing the same or similar work. Which is exactly what we see happening in modern engines, 'less displacement and more work'.


So now review rod bending discussion worrys when using the word "torque" casually which sparked this message length:
1. the use of the words 'torque output' casually as from dynometers does not account for internal energy which affects stress calculation.
2. the radius of the real engine will change force in rod, which will affect connecting rod stress calculation
3. angles change constantly between rod and crankshaft which changes force in the rod which effects stress calculation
4. pressure changes over time which constantly changes force in the rod which affects stress calculation.



For 1 through 4;
1. the more mass (engine/any) parts have the more energy goes into those parts (like rods and wheels, spinning parts) as they accelerate to some velocity, will be more difficult to slow down. Force applied to change direction or velocity will need to be greater when considering more massive objects.
2. as radius goes up, less pressure can be used to generate a similar force, likewise the same magnitude of pressure/area results with more force.
3. near 45* is best transfer but it never stays at just one number, rather there is a rate of change of the angle theta term called "d theta".
4. "d pressure" dp term in an integral to get the area of of a pressure curve as it proceeds along in time while piston is in motion. Torque is provided by a constantly changing pressure over time during power stroke and so rod stress is also changing constantly.


and the combustion reaction only persists through a small portion of total rotations. The 'bang' part lasts breifly and during that time pressure (and connecting rod instantaneous stress due to that pressure) changes constantly. That is why I called torque 'instantaneous', since we need to measure it at a specific moment, for our purposes we want to see the maximum pressure moment, and also the graph of it changing. An engine rotates circles but only provides combustion for a small part of that 720* cycle. The rest of the engine torque is negative, the engine is decelerating. All of this is being factored into a turning engine. an example:

Lets try to measure the torque of a single cylinder engine.
If the engine's box label says "5 horsepower engine" you stick it into this equation at suggested rpm and work backwards and find that it 'has'
8.7 ft*lbs of torque at 3,000rpm. So for a 4-stroke engine, 1,500 times every minute or 25 times per second the flames explosion pushes down on top of the piston for a breif moment, and the rest of the cycle it slows down and rpm drops. The engine decelerates between combustion events. Under typical circumstances there seems to be a deceleration of the engine internals between burn events and I would think especially so for a 1-cylinder engine, giving credit to engines with more cylinders which it looks like would possible (guessing) to eliminate the subtle discontinuities between burns at least with the crankshaft turning rate of change being always positive. Since we are trying to use it in terms of distance (but circumference instead of straight line) it is strange to think of something speeding up and slowing down 25 times per second in a circle and measuring that distance around one that has a 1-foot radius. In reality we just see the thing spinning sometimes so fast it can't be seen. The torque 'number' that we calculated is only the part of the story because in reality the engine could have gone farther if we neglect friction and worldly constraints, if it doesn't decelerate between pulses. And here is an interesting connection, imagine if the engine isn't accelerating but it's radius is less than 1 foot. The torque output measured this way will be less than the actual engine produced and felt at in it's rods (rod stress). The engine we just looked at "5 horsepower 1-cylinder" made an engine with a 1 foot radius rolling move 8.7 lbs 1 foot or 1 pound 8.7 feet. It doesn't say anything about what the actual engine experienced in force/stress values on it's rods or how much torque is actually applied to that engine's crankshaft because its real radius is like an inch or two, or something ridiculously small compared to 1 foot. What I am trying to drive at here is that the rods of the engine in question need to be specified for stress analysis, along with it's materials and so forth to get anything actually calculated, and that you need to refer to the actual engine crankshaft radius to find real torque values, in an instantaneous (at the exact moment of highest pressure or at other times if desired) method, because engine torque in the positive sense only occurs during a very small portion of the total 720* cycle. Furthermore, the internal energy stored in the rotating parts will keep it turning in the event there is a misfire, albeit decelerating more. The internal energy is also a factor as the piston is ascending, since any force from above during this time would apply torque in the reverse-sense direction of the engine and slow it down. Hot spot detonation, high temperature pre-ignition circumstances, or early timed spark all create pressure while the piston is going up and the crankshaft is turning and still turning because of internal energy components. Some of the stresses in the rods and pistons are highest during these times because I expect expanding gasses in a shrinking volume surely will result with disastrous pressure magnitudes. Many rod stress related failures and rod bearing failures are due to stresses which occur before the piston reaches TDC compression, and also I guess seen is internal energy at TDC into the vacuum of a cylinder breaks the rod (circle track deceleration example is common).
random poor quality quote for that which I am not familiar with but read a couple times
"That’s why rod failures most often occur when the driver suddenly lets up on the throttle at the end of a run down a drag strip or when a circle track car exits the straightaway and enters a turn. "
https://www.enginebuildermag.com/201...-rod-failures/


Now that I separated the 'pre' and 'during' TDC stresses from pressure due to combustion occurring when it was 'supposed to'.
When is it 'supposed to'? When the angle of the rod is in a position to utilize the pressure from descending piston area then pressure can be allowed to push down during that entire period. However having a reaction going for the whole thing is kind of rare and unseen in gasoline application (i.e. diesel does it better). The combustion reaction is quite fast and can be faster than the engine can move. If the engine rpm is very low then the engine is rotating slowly- and the fuel is too fast (Notice how fuel quality is the real problem) then the pressure can rise rapidly- as if the piston were not even moving fast- and something will usually break due to this intense pressure "the head gasket's job to break first". If an engine can produce enough pressure in this instant and the HG doesn't blow- its the HG's fault that the rod is bent or piston fractured. It is unlikely to damage a piston/rod if the HG is tailored to suite the strength of the piston/rod so rod/piston failures due to intense pressure is seen as an engine calibration fail or driver error(will get to this later), not a parts failure and certainly avoidable under modest circumstances. Again, theory is most rod failures are during-pre TDC and unrelated to useable torque based failures. The reason I say usable is because that pressure might have actually been fine- if the rod, piston, cylinder, HG, etc.. were all designed to operate at that level. We all know what a cylinder 'O-ring job' more or less is right? Now you can manage very high pressure without HG failure. So hopefully the rod is designed with that in mind. So far I haven't touched on a single problem with torque and low rpm. If the fuel is too fast, its a fuel issue. If the HG didn't blow before the rod failed, its a setup issue. If you use an insane HG then you use the right rods and so forth. And if driver error...


So lets get real for a minute. Gasoline is great most of the time for budget daily drivers due to lowest cost. It is very temperature and pressure sensitive, and I notice they never mention the temperatures at which compression ratios are tested when quoting octane. Reactions usually procede when collisions between molecules occur with sufficient force in the right orientations. This force is from velocity and frequency, and sometimes incidence plays a role (hinderance of other molecules or having to hit a specific spot of a molecule to react) which is related to the number of total molecules per unit volume (mass density since we usually assume all gas molecules are the same size in generic equations). This also helps explain why richer than stoichiometric a/f ratios seem to give more torque, because more molecules are able to find partners with which to react during the short time the piston/rod is in a more suitable position to transfer torque to crankshaft. Since octane is usually related to compression ratio, and most of us have seen naturally aspirated engines ping on 93, the idea of doubling or tripling the atmosphere on 93 seems laughable. And yet it is done frequently, although these days with help from reaction cooling/slowing additives as water/methanol and intercooling. We can use gasoline, just need to be aware of it's shortcomings and raise the octane as necessary or cool it down to limit EGT for safe operation continuously.

Kingtal0n

Courtesy demands I now compare high torque E85/methanol/cold air fuel situation with gasoline, to ask the question, "but what about engines at low rpm with very cool mixtures(slow enough not to spike) like E85 make a lot of torque and seem to bend a rod?". Two paths common

1. timing with E85 can be 'over-advanced' without the user being aware or knock sensors showing knock or dynometers getting very jagged lines (because it isn't knocking). Too early, another 'near TDC' pressure becomes a powerful vertical component force before the rod is at a good angle to transfer the vertical component as torque (the perpendicular distance is too short, the spark was too early and there is too much pressure too early) so instead the rod becomes loaded vertically into the journal wears/ruins the bearing and more stress in the rod because now it has both a powerful vertical component from above (the reaction force of crank lever pushing back against the loading rod, deforming bearing shell and wearing journal bearing), the almost equal and opposite force from the crankshaft below is compressing the rod(more than it would have been if the rod was at a proper angle, since the torque would be greater thus engine would rotate faster which would make more volume in the cylinder faster, controlling pressure increase rate due to combustion and preventing the spike behavior /thread), again this is 'around TDC effect' we already covered but for slow moving(high quality) poorly timed fuels, since the force is a gradual loading it could be bending the rod as pressure will now be higher than it would have been if the angle of the engine were more properly positioned and still providing similar torque (slightly less but due to spiking cylinder pressure can create spikes in torque curve which cause the overall max recorded values increase). In other words, #1 says an early, improperly timed ignition BTDC event can lead to bending rods with no engine knock present and adequate fuel quality at low/insignificant engine torque.
Ex 2
An engine you are familiar with is spinning 7,000 rpm and we measure the torque using engine dynometer.
Lets say the torque is 500ft*lbs, and cylinder pressure peak is measured 1500psi.
So let me ask a question now, what is the cylinder pressure at 6,000rpm?
For torque to be the same 500ft*lbs, the cylinder pressure needs to be basically the same. the only thing that could 'help' is if there was less pressure but it was spread out over more of the cycle, but there is only so much cycle to spread the force over and gasoline type fuels usually finish fast. So now we have just stumbled upon the key to tuning the spark timing of a combustion engine. Have you ever noticed that an engine seems to run very similar within a wide range of timing values for regular driving? If the spark is sent too early- just a little too early- what does that mean for engine output and parts stress? Lets try to imagine the brink of this occurring, you are advancing the timing sooner and sooner until what- its just too much, for... ... well, it depends on when this is happening. If you are cruising on the highway then the cylinder is mostly a vacuum to begin with. Molecular interactions are slower unless its very hot and especially if its lean. Having a spark occur a hair 'too soon' probably wouldn't even be noticed because the only real definition for 'slightly too soon' is that some of the energy that could have went to torque at the rod was unable to transmit that force fully as the rod wasn't quite into position to accept it. There are two components. The component of the force which is always lost due to the angles, and the component of force which was pressure that is unable to transmit force to the crankshaft due to it being set too early. As you set spark earlier and earlier you run into more "pre-TDC" and "during TDC" rod stress scenarios which I have already covered, and are unrelated to "high torque at low rpm" since any pre-tdc and during tdc rod stresses are not contributing torque to the engine crankshaft, the force is instead of being applied to the crankshaft is being transmit improperly into the rod and causing excess stress and strain there.

new situation; Proper ignition timing (not too early) with a sufficient angle(no poorly transmitted force) with a cold/slow enough reaction to not cause a pressure spike (perfect combustion) which is still capable of bending a rod (but not blowing the HG). This is what everyone says 'low rpm torque will bend the rod but high rpm torque will not' as somewhat of a myth in this situation and we will see why in a couple seconds. In the above example of the engine at 7k and 6k we compared combustion pressure and found it to be essentially the same at 6k and 7k in order to make identical torque. In other words, to change the amount of torque relative some given(constant) combustion pressure curve area, you need to do one of two things:
-change piston area
-change crankshaft length

And neither of these things happens at low rpm or high rpm, the crank and piston stay the same size (for our purposes). Therefore torque and combustion pressure will also match, i.e. if I see 500ft*lbs of torque with 100% cylinder fill at 7,000rpm,
I will also see 500ft*lbs of torque at 5,000rpm with the same exact cylinder fill 100%.
That also means at 3,000rpm I will see 500ft*lbs of torque at 100% cylinder fill.
Notice torque never changes and combustion pressure is staying constant (no additional rod stress due to combustion at any rpm while holding torque constant). The only thing that changes from RPM to RPM is power, and rod stress due to RPM (internal energy, having to change directions at 7k rpm is more difficult) which means rod stress actually goes up with increasing RPM while torque stays constant. And this should seem logical

review
the amount of force required to generate 500ft*lbs of torque at 7,000rpm is identical to the force required at 500ft*lbs and 3,000rpm
To put it another way, if 100% cylinder fill gives 500ft*lbs at 7,000rpm then 100% cylinder fill will give the same torque at any rpm, and the force generated (total area under force curve) will be the same.
And, as RPM goes up, the rod experiences more stress due to internal energy.

key feature:
Knowing this- you should be able to tell that there is no way a rod can bend at 3,000rpm with the same forces/cylinder fills as 7,000rpm. The one simple fact that torque = cylinder fill should de-myth this 'low rpm rod bending myth' i.e. if the engine can support that much cylinder fill at 7,000rpm then the stress should be even less in the rod at 3,000rpm or lower.


exceed the rod's designed limit for force its ur own fault
Lets look at how we can maybe bend a rod 'at low rpm with high torque' with the fuel and timing (tuning) and ultimately the pressure area under of the curve of volumetric displacement (it must be done as the piston moves).
torque output only occurs during a small portion of 720* rotation, this time can only happen if the piston is moving, and recall that during the power stroke, the total torque is a combined area under the curve of pressure function on top of the piston as it moves. If the reaction rate is held constant for combustion, then the slower we allow the piston to move (the more heavier the drivetrain (acceleration) or the slower rate the engine is turning (velocity)) the higher the pressure will climb in the cylinder since we are holding reaction rate constant, even while torque output is the same. if we continue slowing the piston down while holding the reaction rate constant we would see the pressure peak higher, until point at which it spikes high enough to pop the head gasket or blow a hole in the piston, or bend a rod. these situations are synonymous; they all pertain to pressure spikes within the chamber which should never happen. In other words, pressure spikes are tuning issues or driver error. In gasoline applications you see this all the time when drivers 'lug the engine' of regular cars, causing excess heat, pressure spikes, at low rpm, which could have been avoided by using higher quality fuel (fuel issue) or by not trying to accelerate the rotating assembly from such a low rpm (shift to better gear, driver error). Nothing to do with 'too much torque at low rpm'. In high performance world, we would say the same thing, either the fuel wasn't able to cope with the situation and it exploded violently (pressure spike due to fuel quality) or the driver was using the wrong gear (ex. 5th gear at 1,200rpm and trying to make a lot of combustion pressure with slow moving parts = pressure spikes) just like any other engine. The key here is that pressure spikes are occurring which damages parts, not just rods but everything, and pressure spikes are unwanted and occur in such a short time that often the energy they release is dissipated as heat instead of used to drive the crankshaft, therefore having pressure spikes present will reduce torque output and cause damage if they are significant. Instead of normal torque application, the engine will "knock" sometimes audibly and torque output will suffer. That means if we are holding torque output constant (as above example 500ft*lbs was constant for each rpm and cylinder fill) that this scenario does not apply, i.e. if we have 500ft*lbs of torque at 7,000rpm AND 500ft*lbs of torque at 2,000rpm, then there must not be any pressure spikes because they would reduce torque output and damage parts, and raise temperatures while wasting energy i.e. the engine wouldn't be running properly and torque output would suffer. A note worth mentioning is that it is possible to have a 'little too much pressure at a specific moment in time' which may be considered a pressure spike, and will also increase torque output if the spike occurs at the right moment; This is a timing issue and it refers back to the "during TDC" stresses we already talked about, it will cause little spikes in the torque curve of an engine output and raise the max torque an engine appears to make on paper. It doesn't blow the engine apart (at that temperature) and may not blow the headgasket but ultimately the pressure being so high, so early will cause a failure (most likely rod bearing failure)

quick Review
-We eliminated pressure spikes (poor fuel (too much compression or too hot for the fuel) or poor timing(too early), with or without too low of an rpm) as a cause of torque related parts failure, in other words torque output is unrelated to failure due to pressure spikes for variety of reasons a.) because in the event of pressure spike, total (combined, integrated torque per pressure curve) torque is decreased, not increased, and we are looking for highest torque/high cylinder fill bending rods which pressure spikes do not support and b.) pressure spikes are always unwanted no matter what their affect on torque output and c.) engines with very low torque output (ex. 120ft*lbs from a 4-cylinder) are just as susceptible to pressure spikes related parts failure (failure happens with low torque, torque isn't the reason for pressure spike related failure). And d.) pressure spikes happen "suddenly" and are more associated with 'sudden fragmentation' such as piston chunks, wristpin/cap failures or brittle behaviors, rather than the kind of force required to bend a rod which is a slower loading, plastic behavior of ductile materials that typically takes more 'time'.
-We showed that 100% cylinder fill at high rpm causes more stress than 100% cylinder fill at low rpm to the rods due to internal energy while torque output is held constant (say, 500ft*lbs). In other words 500ft*lbs of torque at 7k is more stress on the rod than 500ft*lbs at 3k, so the rod is 'safer' at 3k and less likely to plastic if pressure is held identical.
-We acknowledged lower rpms and lower rate of change of rotation(dtheta) require slower fuel reaction behavior, which prioritizes fuel quality and driver gear selection ('lug' an engine) as the culprit that causes damaging pressure spikes at low rpm, and not related to torque output (4-cylinder example being 'lugged' still experiences engine damage at low torque output)


To this point I merely divulged each scenario for piston motion, how it can be damaged going up, going down, pressure spikes, torque and cylinder fill, forces within the rod we can 'see', fuel quality plays a role, etc... We all know that enough force can bend a rod; other situations where torque output is below the limit (the rod should be able to handle 'this much torque') or even zero torque from the engine, and the rod can still bend due to internal forces. A basic ex is hydrolock, bends a rod and the engine stops. If the rotating assy is light enough, a hydrolock will spin the engine backwards instead of bending the rod when it encounters a sudden stop. This relationship of internal energy (rotating mass) to being able to bend a rod is one of the points I keep going back to because it is essential for calculating forces in rods and an unknown when calculating engine torque through the use of engine/chassis dynometers. In other words, knowing instantaneous torque applied to the crankshaft is a useless quantity by itself; we also need the rotating internal energy, rod angle, crankshaft stroke, & forces being applied to the rod, to be able to tell if it will bend, break, spin in reverse, etc...

But we never really discussed 'what to do' or 'how this applies' to our turbo selection, how you can be confident that you do not buy 'too small of a turbo'. That will be the final section and hopefully throw some pictures in here.

------section 3 turbo selection
summary: "if this was just one sentence liner"
Choose the compressor based on tech and flow rate criterion, choose the turbine based on application: daily drivers in cold weather can utilize smaller turbines will still have reasonable torque at part throttle cold, whereas exhaust gas fully expanded in racing is utilized more efficiently when the turbine is pre-warmed with energy, thus insulation and ultimate expanded exhaust gas volume at anticipated output situations (how much power/heat and for how long, a boat might run an hour at 6k vs a car you start and drive to work every day) means that larger turbines are favorably utilized in situations where pre-warming to "ready temperature" is a givenand ultimate exhaust gas volume is recognized (perhaps the opposite of a cold started daily driver).
I'll do 'in other words' for fun: turbines have conditions where they operate most efficiently, so size the turbine to match the conditions you intend to encounter most or as per application specific.


no need to post thread topic done / cleaning up for historical reference