Dumb Idea? Engine Building Strategy.

What you've touched on is one reason why aftermarket tunes add so much more torque than hp - the OEMs are already using boost control to design to a torque limit. Efficiency also largely comes down to throttle angle - the closer to WOT, the more efficient (in the thermal sense, not the mpg sense). So, by allowing low boost, you run a wider throttle opening and gain efficiency. In fact, boost is brought into the engine management routine so some cars will go WOT and then manage output with boost control rather than throttle angle. 

Variable valve timing really is the key to all this though. You get very inefficient boost if you are just trying to cram more air in to compensate for not wanting the ideal cam phases. With boost comes heat, and intake heat is very bad. The better your valve timing, the less stress on the engine package for the same level of torque. 

The biggest downside to this low-torque strategy is the consumer. They don't want rpm, they want torque. I am perfectly fine downshifting to get gearing to do its job, but the general buying public wants what the OEMs call "torque reserve". There is a positive buying response when you only have to dip your toe a little to take off without the engine having to shift. Rarely do buyers actually hit peak hp numbers just because they don't want to spin to redline. 

I would argue that variable valve timing is simpler and easier to implement than a full turbo system.

Also, as I was reading you initial thoughts, I was thinking about big cubic inch Cadillac v8s.  Low rpm, fairly light for what they do.

Two secrets I can think of:  small displacement (shifts the peaks up in the rpm band and limits torque) and if you want boost, just oversize the turbo a bit so it doesn't come on until higher in the RPMs.  That was a trick Porsche used on air-cooled turbos to prevent overheating.  The turbos didn't even really come alive until 4000.

I might look to something like a V-twin motorcycle engine.  The output you're wanting looks a lot like my Kawasaki Vulcan 1500's output.  The downside to smaller cams is that (to a certain extent) you end up potentially making more torque at a lower RPM... hence my suggestion for smaller displacement.

Hp = (Tq x RPM)/5252

You can't escape it.

Prius engines are like 13.4:1 static compression (which is dropped by the Atkinson cams to 9.4:1 or whatever). Switch to the otto cycle cams, add ITBs, fuel, and tune, and you get 150whp:

Similar concepts work with the larger displacement Toyota hybrid engines as well. 250-ish from a naturally aspirated 2.5L isn't too shabby:

cyow5 said:

What you've touched on is one reason why aftermarket tunes add so much more torque than hp - the OEMs are already using boost control to design to a torque limit.

The other factor here is that the compressor side of a turbo has flow (and thus power) limits, and OEMs usually size them without much margin above stock horsepower in order to optimize spool.  So the aftermarket tunes add a bunch more torque because that's really all they can do without changing hardware.

Even the aftermarket can shape boost curves without all that much trouble. We used the VVT to adjust the boost delivery on our ND turbo and work around various limitations. In our case, we were limited on top end power by fuel delivery but we got that turbo to deliver bags of torque down low. This made the car much faster than the power rating would have you expect. If we were trying to get away with a lighter transmission, we could change that strategy. 

But the consumer does not like high revving engines unless it's for sporting purposes.

As for centrifuguals, they feel weird to drive on the street - you could probably get around that by using an automatic and some throttle plate management. They also sound like they're self-destructing at idle which is a little harder to deal with for an appliance car.

WRT emissions- other than the first 15-30 seconds, I would not really do anything for the engine to "lower" emissions. We know enough about catalyst control that the engine doesn't need to really do anything.  I would (and have) told the engine designers that they should do as much as they can to use the excess HC and CO to make more power, though. Look at it as free efficiency improvement. 

I know when I went shopping for transaxles for rear and mid engine cars there are just not a lot out there in smaller sizes.

Hewland and Quaife  will make you monsters that can handle huge power, but were too large and heavy for our needs. Sadev made gearboxes that fit our size requirements, but they pleaded with us "No more than 300nm please"

That is likely the whole reasoning behind keeping the torque numbers low.

Centrifugal chargers are like having big, laggy turbos without all the horsepower. Don't forget the drive losses as you aren't getting some for free from the exhaust heat/pressure. You can still size a turbo mildly and have the boost come in flat/slow to avoid engine strength issues. In general rpm kills things way faster than torque but depends on engine design. 

codrus (Forum Supporter) said:

cyow5 said:

What you've touched on is one reason why aftermarket tunes add so much more torque than hp - the OEMs are already using boost control to design to a torque limit.

The other factor here is that the compressor side of a turbo has flow (and thus power) limits, and OEMs usually size them without much margin above stock horsepower in order to optimize spool.  So the aftermarket tunes add a bunch more torque because that's really all they can do without changing hardware.

 

This was why I was thinking a centrifugal blower. You could have a large impeller and gear the supercharger so that it doesn't come on until the higher rpm ranges. Such a blower would do almost nothing at low RPM and then have tons of flow up high. A clutch would seem to cut out the parasitic drag when doodling around as well. 

Trent said:

I know when I went shopping for transaxles for rear and mid engine cars there are just not a lot out there in smaller sizes.

Hewland and Quaife  will make you monsters that can handle huge power, but were too large and heavy for our needs. Sadev made gearboxes that fit our size requirements, but they pleaded with us "No more than 300nm please"

 

That is likely the whole reasoning behind keeping the torque numbers low.

It's exactly the reason the Sendy Club had. 

Keith Tanner said:

Even the aftermarket can shape boost curves without all that much trouble. We used the VVT to adjust the boost delivery on our ND turbo and work around various limitations. In our case, we were limited on top end power by fuel delivery but we got that turbo to deliver bags of torque down low. This made the car much faster than the power rating would have you expect. If we were trying to get away with a lighter transmission, we could change that strategy. 

But the consumer does not like high revving engines unless it's for sporting purposes.

As for centrifuguals, they feel weird to drive on the street - you could probably get around that by using an automatic and some throttle plate management. They also sound like they're self-destructing at idle which is a little harder to deal with for an appliance car.

 

Interesting. I've driven some centrifugal LS Camaros and to me they just felt like big engines with VTEC. I didn't notice the idle issue either. 

Curtis73 (Forum Supporter) said:

Two secrets I can think of:  small displacement (shifts the peaks up in the rpm band and limits torque) and if you want boost, just oversize the turbo a bit so it doesn't come on until higher in the RPMs.  That was a trick Porsche used on air-cooled turbos to prevent overheating.  The turbos didn't even really come alive until 4000.

I might look to something like a V-twin motorcycle engine.  The output you're wanting looks a lot like my Kawasaki Vulcan 1500's output.  The downside to smaller cams is that (to a certain extent) you end up potentially making more torque at a lower RPM... hence my suggestion for smaller displacement.

This thought exercise wasn't really connected to anything I own or am working on. I actually got the idea from playing with Automation's (not very realistic) engine simulator. When I want a sporty, 200-ish horsepower mild hatch in that game that gets great mpg and can be built for cheap, a 1.5-ish liter pushrod four cylinder with very mild cam (great idle, good fuel economy, dirt cheap) + forged and rpm friendly bottom end (for survivability) + clutched centrifugal blower that does nothing under 3000 rpm but ramps up to 15 psi at 9000 rpm (for keeping the torque curve extending out to infinity) is kind of a cheat code. I'm not saying the game is realistic, btw, but I am curious. 

I would think the Porsche approach has the following downsides relatively speaking. 
1. Even at 4k rpm, it's going to lag. 
2. The exhaust flow is going to depend on cam timing, so the desired ultra mild cams will just never spool that big turbo.
3. You can't "turn it off" from the standpoint of exhaust backpressure when you're just putzing. Clutching and bypassing a centrifugal blower is quite easy. 

cyow5 said:

Variable valve timing really is the key to all this though. You get very inefficient boost if you are just trying to cram more air in to compensate for not wanting the ideal cam phases. With boost comes heat, and intake heat is very bad. The better your valve timing, the less stress on the engine package for the same level of torque. 

This is fascinating. How much does this extra heat happen? 

In reply to DaewooOfDeath :

How much heat or how does it happen? Your wording is strange. The how is much easier to answer though. Any time you compress air quickly, it heats itself up much faster than it can heat up what compressed it. This is referred to as "adiabatic", and there are simple equations for telling you how much heat is generated. This is why we need intercoolers when running any appreciable amount of boost. How much heat depends on where you are on the compressor efficiency table. This is what determines the limit Codrus mentioned above. At some point, the extra heat negates some of the extra power, and OEMs tend to draw a more conservative line than aftermarket tuners because the OEMs are aiming for power on the nth pull and not just the 1st. 

When looking at efficiency, you need to basically audit where all the heat is going. Heat lost to the intercooler is bad. Heat lost to the exhaust is bad. Turbos are good when they reduce the heat lost to the exhaust while losing little heat to the intercooler. There's your efficiency sweetspot. VVT helps a ton here by enabling more power with less boost (and therefore less heat). In fact, a supercharged car can see a drop in boost when running better cam timing despite gaining power. (I've seen threads online like "why did my boost drop?")

In reply to DaewooOfDeath :

On a blower or turbo engine equipped with blow-through carburetors, the air-fuel mixture can lower the intake air temp over 100 degrees Fahrenheit.

In general terms, intake heat is wasted energy, and anything you do to reduce it improves the efficiency of the system. 

In reply to DaewooOfDeath :

Why pushrod?  All that does is add stuff to the valvetrain which limits how fast the engine spins. 
 

For a V engine, it does simplify enough to justify that, since you eliminate a camshaft. But for an I engine, it just is extra stuff. 

In reply to DaewooOfDeath :

I have been enjoying working through this thought experiment, as this runs tangentially to numerous thoughts I've had, posted about, and continue to have.

Prius doesn't just use 'mild little camshafts' though, it uses a modified Atkinson cycle which trades power for efficiency. Adding forced induction basically turns that into a Miller cycle. Which is great, but doesn't seem to be what you're looking for either. It may not have VTEC (does have VCT though) or high flow ports, but are those really any more 'fancy' than forced induction?

If you're just talking about taking a ~1.5L economy engine, beefing up a few parts, and adding forced induction, pretty much every manufacturer has done that now and it works great. The big difference I can see from what you seem to be poking at is that it also eliminates the need to rev, which is actually a good thing for economical power production.

Running at 8k rpm is much less efficient than 4k rpm. For starters, there is a lot more friction. Some of this comes from things like the stiffer valve springs needed to rev that high too. Based on the BSFC and thermal efficiency charts I've seen, all seeming to peak in the 3k-4kish range, it seems likely that overall efficiency in general is highest at these speeds too... Which would make sense why economy engines are thus cammed for that range too.

So while you may be able to use lighter drivetrain components with the lower torque at higher rpm, there are even greater efficiency losses to be had up there as well. Ever seen the Top Gear where they run a Prius flat out vs an E90 M3 (at the Prius speed) in a fuel economy test? Spoiler: The M3 wins.

At low throttle and low rpm, the smallest and mildest engine wins the economy contest. However, as power needs increase, this doesn't hold true. If you have to run an engine above ~4k rpm to get the desired power, you're generally losing efficiency vs a more (peak) powerful engine running the same power in that sub-4k rpm range.

On to the forced induction. What you described for equal torque across the rev range is not something that a supercharger would be used to do. This is because even the smallest cammed naturally aspirated engines make peak torque in the mid-range rpms. So that would be your point of lowest boost pressure. You'd need greater boost both above AND below that point, with the more boost needed the further you are away. That's probably why ridiculously (artificially) flattened torque curves didn't start appearing until the advent of more advanced computer controlled turbo systems.

So now what you need is a turbo that can make significant boost at very low rpm, which means it will be less efficient on that engine at very high rpm. This is where you run into the excessive heating of the intake charge to maintain the desired output.

DaewooOfDeath said:

3. You can't "turn it off" from the standpoint of exhaust backpressure when you're just putzing. Clutching and bypassing a centrifugal blower is quite easy. 

Sure you can.  The turbine doesn't present much much backpressure at low loads, and even if you felt that that was too high you could use an electrically-operated wastegate to let the exhaust bypass it.

From an OEM's perspective, turbos are better than centrifugal blowers in basically every way.  They're the same efficient compressor geometry, but decoupling the speeds of the compressor wheel and the engine lets you make useful boost over a much broader range of RPMs.  They're also more efficient under boost because backpressure hurts your power less than parasitic drag.  Centrifgual blowers are an aftermarket thing -- they basically only exist because they're cheap and easy to install on a pre-existing car.  The same isn't true for OEMs because it doesn't cost them any more to make a turbo manifold than a normal manifold, for example.

I think the problem is with your original premise about low torque being desirable because drivetrains capable of high torque are too heavy.  That's a concern for someone building a one-off Pikes Peak race car like the sendy club because they are limited to using whatever parts are available in the market right now but OEMs design new transmissions all the time, they aren't restricted to what's on the shelf.

DaewooOfDeath said:

Keith Tanner said:

Even the aftermarket can shape boost curves without all that much trouble. We used the VVT to adjust the boost delivery on our ND turbo and work around various limitations. In our case, we were limited on top end power by fuel delivery but we got that turbo to deliver bags of torque down low. This made the car much faster than the power rating would have you expect. If we were trying to get away with a lighter transmission, we could change that strategy. 

But the consumer does not like high revving engines unless it's for sporting purposes.

As for centrifuguals, they feel weird to drive on the street - you could probably get around that by using an automatic and some throttle plate management. They also sound like they're self-destructing at idle which is a little harder to deal with for an appliance car.

 

Interesting. I've driven some centrifugal LS Camaros and to me they just felt like big engines with VTEC. I didn't notice the idle issue either. 

It may depend on the brand of centrifugal, but every one I've experienced (from multiple vendors) has made rattling noises at idle. If they're attached to a barely muffled LS, you may not hear it. On a four cylinder, it sounds like something is terribly wrong.

They feel sorta like good naturally aspirated engines with a constantly climbing power curve. Boost is directly related to engine RPM, so power increases with engine speed. But because the boost is climbing, you get this odd behavior where a steady throttle position gives you more and more and more power in a way that positive displacement supers or turbos do not. You have to learn your way around it, to be backing off the throttle as you're moving through an intersection for example. Easy enough to deal with with modern engine management and DBW, I suppose. On track, the fact that power falls off quickly if engine speed drops makes them a bit tricky to drive quickly. Get a bit off the boil because you're dealing with traffic and the car goes limp. Don't have that problem with a good turbo or a positive displacement SC.

I know there's a group of track junkies that think they're the best form of aftermarket forced induction for track use because they run a little cooler than turbos and they mostly are looking to "win" an HPDE, but I've never been impressed with them on the street, on the track or at idle. I'm not sure there's been a production vehicle with a centrifugal for nearly a century.

In reply to Keith Tanner :

If the rattle comes from the geartrain in the blower, I'd expect the rattle to be less significant on smoother idling engines.  So milder cam, more cylinders, higher idle speed, etc. should all lead to less rattle.  But a low idling 4 cylinder could potentially produce a lot of rattle. 

cyow5 said:

In reply to DaewooOfDeath :

How much heat or how does it happen? Your wording is strange. The how is much easier to answer though. Any time you compress air quickly, it heats itself up much faster than it can heat up what compressed it. This is referred to as "adiabatic", and there are simple equations for telling you how much heat is generated. This is why we need intercoolers when running any appreciable amount of boost. How much heat depends on where you are on the compressor efficiency table. This is what determines the limit Codrus mentioned above. At some point, the extra heat negates some of the extra power, and OEMs tend to draw a more conservative line than aftermarket tuners because the OEMs are aiming for power on the nth pull and not just the 1st. 

When looking at efficiency, you need to basically audit where all the heat is going. Heat lost to the intercooler is bad. Heat lost to the exhaust is bad. Turbos are good when they reduce the heat lost to the exhaust while losing little heat to the intercooler. There's your efficiency sweetspot. VVT helps a ton here by enabling more power with less boost (and therefore less heat). In fact, a supercharged car can see a drop in boost when running better cam timing despite gaining power. (I've seen threads online like "why did my boost drop?")

I understand this, and I am familiar with "good" boost loss when you put better heads or a bigger cam on an engine without changing the supercharger (or even turbo, depending on the flow limits).

What I'm wondering is if we tune the supercharger to be at its efficiency sweetspot at, for example, 12 psi, does it care from an abiatic standpoint if that 12 psi is with super mild cam timing or if it's with ideal cam timing?

alfadriver said:

In reply to DaewooOfDeath :

Why pushrod?  All that does is add stuff to the valvetrain which limits how fast the engine spins. 
 

For a V engine, it does simplify enough to justify that, since you eliminate a camshaft. But for an I engine, it just is extra stuff. 

This is very meta and only applicable to the game, but I usually use significantly undersquare boxer fours to make the engine as short as possible, then mount it longitudinally in front of the front axle. This is the cheapest way, in game, to make one chassis easily switchable from FWD to AWD.