A few years ago I found a 5 speed and matching R200 rear from an '81 280ZX for sale on the internet. This is widely considered to be the "hot" setup for an S30 with a few more horses than stock, and I snapped it up and added it to my collection of parts. When I started planning my L28 build, I got around to thinking about gearing and looked up the ZX ratios and did some figuring, and was a little surprised - the ZX gears seemed like a poor fit for what I was after. It was time to do some math...
Saturday, November 30, 2024
Wednesday, November 27, 2024
More fuzzy subjects: Camshafts
If you're building a hot L6 it has to have a performance cam. Going beyond cool factor, the stock cam is a pretty mild piece that is going to limit what you can get out of the motor.
Without a modern computer-controlled camshaft system such as Honda's VTEC, choosing a cam (and setting the cam timing) is a trade off between low-RPM drivability and high-RPM power; and it always seems that the choices are either too big or too small. Fortunately there is no shortage of information (although its not always the information that you need). Lets start by looking at what all the numbers mean.
Without a modern computer-controlled camshaft system such as Honda's VTEC, choosing a cam (and setting the cam timing) is a trade off between low-RPM drivability and high-RPM power; and it always seems that the choices are either too big or too small. Fortunately there is no shortage of information (although its not always the information that you need). Lets start by looking at what all the numbers mean.
While doing the final read through of this write up, I realized I've left out lots of practical details behind changing camshafts. It is not a simple bolt on. This is already long enough - I'll do that in the next chapter.
- Lift. The maximum distance that the valve moves off its seat. The higher the valve lifts off the seat, the more air can flow through it every millisecond it is open. The stock Datsun cam has a valve lift around 0.4 inches. Datsun engine builders claim air-flow gains for lifts as high as 0.7 inches, although that much lift is a lot to ask of the L6 valve-train. Most Datsun "street" performance grinds have lifts around 0.500 inches or less, while true racing cams may top 0.6 inches. The way that the Datsun valve-train works, the lift provided directly by the cam lobe is multiplied by the leverage of the rocker-arm. The exact amount of leverage depends on the valve geometry - valve stem height, lash pad thickness, tower shimming, lash setting - so the published valve lift for a cam may not exactly match what you measure with a dial indicator.
- Advertised duration. Duration is the number of crankshaft degrees that the intake or exhaust valve is open. A higher duration cam holds the valve open longer each time, which means compared to a lesser duration cam there is more time for air (or exhaust) to flow through the valve. A complication is that every cam grinder is free to choose the valve lift that they consider "open" for purposes of determining duration; it is typically in the range of .002 to .010 inches of valve lift, which makes it hard to do an apples to apples comparison of two camshafts from different grinders based on advertised duration. Moreover, less than .005 inches of valve lift (about the thickness of 2 sheets of paper) is barely "open", so the duration at very low lift doesn't tell you much.
- 0.050 duration. Back in the 1950s American cam grinder Ed Iskenderian started specifying valve opening and closing points at .050 inches of tappet lift. In an American V8 this was an easy to measure way (with a dial indicator or even just a caliper) to verify that the cam was installed correctly, and many other cam grinders quickly adopted the same standard. Given the zoo of advertised duration specifications, the duration based on the .050 tappet events became a defacto standard for comparing camshafts from different manufacturers. While .050 tappet lift is more than the true "just open" point, it is close enough that the .050 duration gives a pretty good comparison of total valve-open duration.
Note that an over-head-cam (OHC) valve-train as found on the Datsun L6 doesn't have a tappet (or lifter), so the cam grinders typically spec OHC duration at .050 inches of valve lift (you should always pay attention to the spec - and ask the grinder if its not clear - to know how their cam was measured). Because OHC duration is measured at the valve - after the lift is multiplied by the rocker arm - .050 is reached earlier than a similar cam in a pushrod motor would reach .050 of tappet lift. This tends to add a few degrees to the .050 duration of an OHC cam compared to a push rod engine. You see this when you compare the duration of an L6 cam to a small-block Chevy cam: where a L6 street cam might have 230 degrees of .050 duration, a Chevy cam at a similar performance level might be just 220 degrees. - Rocker-ratio. In many engines the movement of the camshaft is multiplied by a simple lever - the rocker-arm - so that the valve moves a larger amount than the lobe lift. Since cam-grinders often produce a large line of camshafts as well as one-offs, they don't install every cam in an engine and measure the valve lift; instead they measure the lobe lift and multiply by the nominal rocker-arm ratio. For the Datsun L6, most cam grinders seem to assume a rocker-ratio of 1.5:1, although people who have actually measured it come up with a number closer to 1.48:1. There is another complication in that the L6 rocker-ratio changes slightly as the lobe contact point moves across the rocker arm contact pad, so that the valve geometry (lash-pad thickness) can affect the maximum rocker ratio. Some engine builders will try to choose lash-pads to bias the lobe contact to the far end of the rocker and so increase the rocker ratio, but the possible gain is tiny.
- 0 Lash. The valve-lash - the amount of clearance in the valve-train - also affects the valve-lift and duration. In normal operation, the valve-lash is set somewhere in the neighborhood of .010 inches; that means the valve won't event start to move until the lobe has reached .010 inches of lift. Because the lash is adjustable and may be set to other than the stock setting for various reasons, the cam grinders just ignore the effect of lash and specify lift and duration at 0-lash (with all the clearance adjusted out of the valve-train), even though the engine should never be run that way. This is all to say that for the Datsun L6, the measured valve lift and .050 duration will almost always be slightly smaller/shorter than what is specified on the cam grinder's web page. Again, pay attention to the small print so you know what the numbers actually mean.
- Lash ramp. The first and last (approx) 15-20 crankshaft degrees of a cam lobe's lift are designed to gently take up the clearance or "lash" in the valve train (so the rocker arm doesn't slam into the valve stem) and gently return the valve to its seat (so the closing valve doesn't hammer the valve seat into the head). Typically these ramps have a very small slope of about 0.0005 inches per crankshaft degree. At the end of the lash ramp is a transition to a much faster opening rate; the .050 valve-lift "check height" used by most cam grinders will typically come 5-10 degrees after the end of the lash-ramp. One implication of this is that just the lash-ramps can account for as much as 40 degrees of the advertised duration of a cam - "duration" where in actual operation the valve has not even moved off the seat (this is why the .050 duration is considered a better indicator of how aggressive a cam really is).
- Lobe center angle (LCA). This is the crankshaft angle - measured from top-dead-center - where the valve reaches maximum lift. Note that this is not necessarily halfway between valve opening and closing - just the point of maximum lift. This angle is typically around 100-120 degrees, and can be different for the intake and exhaust valve. Note that the values of the two LCAs (intake and exhaust) depends on how the camshaft is installed. The camshaft and cam sprocket are designed so that you can shift the cam relative to the crankshaft (advance or retard), which would make one of the LCAs bigger and the other smaller. The published LCAs should be for the camshaft installed in the nominal 0 position.
- Lobe separation angle (LSA), also known as lobe displacement angle (LDA). This is the number of camshaft degrees between the maximum intake and exhaust lift, and it depends only on the camshaft. If you could take a perfect end-on picture of the camshaft, you could measure the LSA with a protractor. Because of the way it is defined, the LSA is always equal to the average of the intake and exhaust LCAs (add the two LCA numbers and divide by 2), so it will also be a number around 100-120 degrees. The LSA - and the cam duration - directly control how much valve-overlap a particular camshaft will produce; all else being equal a smaller LSA will create more overlap and a larger LSA will lead to a later closing intake valve and lower dynamic compression.
- Overlap is the number of crankshaft degrees when the intake and exhaust valve are both open (the number should be qualified as to whether "advertised" or "0.050" valve events were used, but if the number is much bigger than 30 degrees it is probably based on "advertised" lift). Why would we ever want both intake and exhaust valves to be open at the same time? Remember that the exhaust stroke is immediately followed by the intake stroke, and the valves can't open and close instantly, so starting to open the intake valve before the start of the intake stroke allows the valve to be open wider during the actual intake stroke. Likewise, closing the exhaust valve after the end of the exhaust stroke keeps the exhaust valve open wider during the actual exhaust stroke. Since the piston is barely moving near TDC when both valves are open, a little overlap isn't as disruptive as it first seems: at low RPM and part throttle, the burned exhaust is pretty much already gone by TDC and since the piston essentially "stops" at TDC (it is moving much slower than at mid-travel) there is no real pressure or vacuum being created in the cylinder to affect the exhaust flow. At higher (mid-range) RPMs the overlap can be extra beneficial, as the momentum of the out-going exhaust in the exhaust manifold can help pull fresh air-fuel into the cylinder before the intake stroke even starts. Note that the amount of overlap built into the camshaft is fixed, but advancing or retarding the cam can shift when the overlap occurs more towards the end of the exhaust stroke or the beginning of the intake stroke.
- Intake closing point. The crank angle - usually measured from bottom-dead-center - where the intake valve is considered closed (again either "advertised" or the "0.050 lift" point). Where this occurs relative to the piston position has a big effect on the performance of an engine. At high RPM the momentum of the incoming air can continue to fill the cylinder even after the piston has started moving up the cylinder, so keeping the valve open later can help fill the cylinder at high RPM. Making the closing come later - more degrees after BDC - can reduce the amount of air-fuel trapped in the cylinder (by allowing it to be pushed back into the intake manifold) and so reduces the dynamic compression ratio to help reduce detonation, but at low RPM reversion can hurt torque and can cause fuel to be sprayed out of the carbs (especially individual runner systems). The intake closing point at .050 lift is usually a number around 40 degrees ABDC, the advertised number would be around 60 degrees ABDC.
- Advance. The cam gear that drives the cam is designed to allow small angle offsets relative to the crankshaft. If the cam is installed "advanced" all of the valve events - opening and closing points - will come a few crankshaft degrees earlier than specified by the cam grinder. If the cam is offset in the other direction, it is said to be "retarded", and all of the valve events will come later than spec. Advancing or retarding a cam moves the intake-closing point earlier or later, biasing the cam to low or high RPM operation as described above. A lot of street cams have a few degrees of advance built in so that the cam is already advanced when installed in the "0" position (a cam with such built-in advance will have an intake LCA that is smaller than the exhaust LCA).
- Asymmetric cam profile. If you look at a typical picture of a camshaft, the cam lobe looks to have the same (mirror image) profile on the opening and closing side of the lobe. You might think that the valve lift profile matches the shape of the cam lobe but the geometry is more complicated. The "flanks" of the cam lobe don't do much to open the valve; its the tip of the lobe that develops maximum lift. As the valve opens, the lobe does a complicated dance, sliding and rolling its way across the contact area on the cam follower (rocker arm). As mentioned earlier, in the L6 the effective rocker-ratio changes throughout the time the valve is open. To make the valve-lift curve symmetric, the lobe requires an exaggerated "opening" side to account for the lower rocker ratio during the initial part of the valve opening.
The stock Datsun cams are just a little-bit asymmetric to give a nearly symmetric valve lift profile. Some cam grinders go further and design an even more asymmetric lobe to slam the valve open as fast as possible - faster than the valve can close. Remember that when the valve opens there is a solid steel-on-steel linkage pushing the valve, but when the valve closes it is the spring doing the work: no matter how aggressive the closing side of the cam lobe is it can't close the valve faster than than the spring can pull the valve back onto the seat. Such cams generally require a lot of development, so you're most likely to find them from cam grinders that specialized in all-out Datsun racing cams back when the L6 was state of the art (Isky is one of the few still around). If you think about it a bit, an asymmetric cam profile is more important to a small duration cam where the fast opening can take advantage of the limited valve open time. - About the stock Datsun cam. The stock Datsun cams have advertised durations around 248 degrees; Datsun never specified a .050 inch duration for the stock camshafts. Over the years, the Japanese auto industry has settled on a 1mm (.039 inch) valve-lift check-height for camshafts, which makes the 1mm duration spec roughly comparable to the .050 duration (for the same cam, you would expect the 1mm duration spec to be just 3-4 degrees longer than the 0.050 duration). But that is almost certainly not how the stock Datsun L6 cams were specified. A cam with 248 degrees duration at .050 or 1mm lift would be on the high end of the most aggressive street performance cams available, while the stock Datsun cams are pretty docile. People who have measured the stock Datsun cam with a degree wheel and dial-indicator have found valve lift of .420 inches and 0-lash .050 duration around 200-210 degrees - entirely believable numbers for a mild street cam.
The stock cam LSA is 109 degrees, which would put the intake closing point - based on .050 check height - around 29 degrees ABDC. Many off-the-shelf aftermarket cams share the 109 degree LSA, probably to make it easier to regrind the performance profile on a stock cam, although there is nothing magic about that number. - Stage 1, 2, 3, 3/4 Race, etc. Although these terms are thrown around a lot, they are basically marketing terms that don't have specific meanings. When performance cams first became a thing with hot-rodders back in the mid 1950s, things were a lot simpler. There were only 2 or 3 engines that mattered (flathead Ford, small block Chevy, Chrysler "Firedome"). There were also few options: cam grinders might offer just 1 or 2 different grinds for each engine, generally tailored for drag or dirt-oval racing. As that technology filtered down to the street there was demand for a broader range of cam grinds that would be more practical for less than all out race performance. Different cam grinders came up with different terminology: early street grinds were often called 1/2 race cams, and one grinder created the "3/4 race" name for a cam with a "full race" intake lobe and a "1/2 race" exhaust lobe. Other cam grinders came up with "stage 1" , "stage 2, etc. terminology to give an indication of how aggressive a cam was. "Stage 1" generally implied a bolt-in upgrade for the stock cam - a grind with slightly more lift and duration than stock and capable of working with the stock springs and retainers. Stage 2 and 3 cams increased lift and duration still further - typically at the cost of street manners and requiring different springs and possibly other mods to work. Some cam grinders offer "stage 4" and beyond cams, which are generally borderline race grinds - they appeal to the car owner who wants maximum performance and doesn't care how rough the idle may be.
- Secret Sauce. There are legendary cam grinders who were known for making cams that performed better than everyone else's cam with similar lift and duration numbers. Often the exact specs for these cams were "secret" - sometimes the only number published was maximum lift - and a vague description like "extreme top-end". Clearly these cams couldn't be all that different: the basic lobe shape is determined by the mechanics of the valve train. What is possible is that a cam grinder figured out a profile that opened the valves a little faster - without putting the valve spring into oscillation or putting too much extra wear and tear on the lobes and rockers - and took advantage of that to tweak the opening and closing points a few degrees, and so made a few percent more power than their competitors. An extra 10 horsepower at redline is a big deal on a race-track, not so significant for a street driven car.
So how do we put all that together? There are lots of rules of thumb explaining how changes in lift and duration and LCA and LSA affect engine performance, but you will go crazy trying to juggle all these numbers and guidelines. Lets see if we can simplify things a bit. I'll focus on the intake valve timing, since that is where stock engines usually need help.
First lets think about lift. Clearly if the valves are open further they will flow more air, and just increasing lift doesn't change the valve-event timing in ways that hurt low-RPM performance and drivability. But reducing restriction at the valves isn't going to have a big benefit if the intake and exhaust ports and manifolding are still restrictive - more valve lift just moves the flow bottleneck somewhere else - taking advantage of increased lift will require improving the flow capacity of the intake and exhaust systems.
The amount lift can be increased is also limited by how fast the valves can be opened and closed without causing excessive strain and wear on the valve train; adding say 20% more lift doesn't necessarily increase the lift 20% across the whole valve profile because the opening and closing ramps may be at the limit for how fast the valve can be opened or closed. So while more lift generally can't hurt, it doesn't have the potential for a big payoff in a street cam.
Which brings us to duration. More duration will take advantage of the existing flow capacity of the head and manifolding by keeping the valves open longer and opening them wider during the actual intake stroke. We're not talking about big differences in .050 duration between stock (200 degrees) and the "hottest" street cams (245 degrees), but remember that the actual intake and exhaust strokes are only 180 crankshaft degrees. Even small increases beyond 180 degrees will have a fairly big impact on the duration that spills over into the previous and next stroke.
Imagine we add 10 degrees to the stock cam's intake duration. If we add that 10 degrees at the opening end of the valve lobe, we increase the overlap by 10 degrees; adding it at the other end will move the intake closing point 10 degrees. In addition to the improved mid-RPM scavenging, 10 degrees of overlap adds about .050 inches of valve lift early in the intake stroke - good for low RPM power. Delaying the intake closing by 10 degrees can boost high RPM power, but sacrifices volumetric efficiency at low RPM and has the potential of creating flow reversion.
The stock cam and even "mild performance" cams have no significant overlap (the intake and exhaust lash-ramps may overlap, but the valves aren't really open at the same time). An extra 10 degrees of duration that all goes to increasing overlap, and gets the intake valve open wider early in the intake stroke, can make a pretty big difference even with a restrictive intake and exhaust system.
First lets think about lift. Clearly if the valves are open further they will flow more air, and just increasing lift doesn't change the valve-event timing in ways that hurt low-RPM performance and drivability. But reducing restriction at the valves isn't going to have a big benefit if the intake and exhaust ports and manifolding are still restrictive - more valve lift just moves the flow bottleneck somewhere else - taking advantage of increased lift will require improving the flow capacity of the intake and exhaust systems.
The amount lift can be increased is also limited by how fast the valves can be opened and closed without causing excessive strain and wear on the valve train; adding say 20% more lift doesn't necessarily increase the lift 20% across the whole valve profile because the opening and closing ramps may be at the limit for how fast the valve can be opened or closed. So while more lift generally can't hurt, it doesn't have the potential for a big payoff in a street cam.
Which brings us to duration. More duration will take advantage of the existing flow capacity of the head and manifolding by keeping the valves open longer and opening them wider during the actual intake stroke. We're not talking about big differences in .050 duration between stock (200 degrees) and the "hottest" street cams (245 degrees), but remember that the actual intake and exhaust strokes are only 180 crankshaft degrees. Even small increases beyond 180 degrees will have a fairly big impact on the duration that spills over into the previous and next stroke.
Imagine we add 10 degrees to the stock cam's intake duration. If we add that 10 degrees at the opening end of the valve lobe, we increase the overlap by 10 degrees; adding it at the other end will move the intake closing point 10 degrees. In addition to the improved mid-RPM scavenging, 10 degrees of overlap adds about .050 inches of valve lift early in the intake stroke - good for low RPM power. Delaying the intake closing by 10 degrees can boost high RPM power, but sacrifices volumetric efficiency at low RPM and has the potential of creating flow reversion.
The stock cam and even "mild performance" cams have no significant overlap (the intake and exhaust lash-ramps may overlap, but the valves aren't really open at the same time). An extra 10 degrees of duration that all goes to increasing overlap, and gets the intake valve open wider early in the intake stroke, can make a pretty big difference even with a restrictive intake and exhaust system.
Its probably a good time to discuss "too much cam". A lot of old-school hot-rodders learned the hard way that a monster cam in a small block V8 made for an un-drivable beast that woudn't idle and didn't make enough manifold vacuum to power the brake booster. That experience gave rise to lots of advice to not go "too big", or more specifically "to avoid too much overlap" on an aftermarket cam. The problem the hot-rodders ran into is that in a V8 with a single plane intake manifold (8 cylinders all sharing the same plenum), during each 2-crank-rotation cycle of the engine, when any cylinder is going through overlap, another cylinder is right at the beginning of its intake stroke. The intake manifold vacuum generated at idle by the cylinders on their intake strokes can potentially suck exhaust backwards through the cylinder going through overlap and into the intake manifold, which dilutes the air-fuel mix in the manifold and reduces the vacuum that the carburetor relies on to control the idle - and that powers the brakes! A turbo tends to make the overlap problem worse - and the stock turbo intake is a single plane manifold - cams designed for a turbo application typically have a big LSA and non-existent overlap.
The neat thing about a Datsun L6 with dual SU carbs on the stock intake manifolds is that for any reasonable amount of cam duration there is no interaction between one cylinder's overlap and another cylinder's intake stroke as out of the three cylinders fed by a carb no two intake valves are ever open at the same time (that isn't true of the L28's stock fuel-injection system, which has a common-plenum intake manifold feeding all 6 cylinders). Too much overlap can hurt gas mileage and emissions by allowing some of the fresh air-fuel mix to flow out of the exhaust, but for a restrictive head with relatively small valves such as the L6 a small amount of overlap is unlikely to cause drivability problems and will boost mid range performance. The hard question to answer is how much overlap is "small"; common wisdom seems to be less than 20 degrees of 0.050 overlap is OK for a street driven car.
What can contribute to low RPM drivability problems, especially with side-draft carburetors, is a too late closing point for the intake valve; having the intake valve open at low RPMs as the piston starts up the bore can push air-fuel back out the intake port - so called reversion. With an individual-runner manifold and side-draft weber-type carbs there is little room in the manifold to absorb this backflow - the reversion can push air-fuel mix back through the carb and into the air-cleaners. The most common cause of a late intake closing point is a big duration cam with a big LSA - where the big LSA is often spec'd along with the big duration to avoid too much overlap!
Area Under the Lift Curve: Internet cam experts talk about "maximizing the area under the lift curve" - adding up the valve lift at 1 degree increments to get a sort of "total lift" measure for a cam profile. Its not clear how meaningful this number is, and you can't accurately compute it with just a single lift and duration number - you need to know the exact cam profile, and cam grinders are often pretty secretive about this.
But never underestimate a geek with a spreadsheet... I came up with approximate cam profiles to try to get a general idea of how lift and duration affect the area under the curve. And what I found wasn't very surprising: increasing lift 10% increases the area under the curve about 10%, and increasing duration 10% also increases the area-under-the curve about 10%. A couple things were interesting (and I don't know how meaningful any of this is, as there was lots of approximating going on):
Back to the real world. Practically speaking, there are only a few choices for an off-the-shelf street performance cam (many cam grinders will happily make a cam to whatever specs you want, but their standard grinds are going to cover most typical applications). Most cam grinders offer a camshaft with near stock-lift and slightly longer than stock duration, a camshaft with near-stock duration and slightly higher-than stock lift, and a camshaft with both slightly higher than stock lift and duration (there are of course lots of variations within these basic options).What can contribute to low RPM drivability problems, especially with side-draft carburetors, is a too late closing point for the intake valve; having the intake valve open at low RPMs as the piston starts up the bore can push air-fuel back out the intake port - so called reversion. With an individual-runner manifold and side-draft weber-type carbs there is little room in the manifold to absorb this backflow - the reversion can push air-fuel mix back through the carb and into the air-cleaners. The most common cause of a late intake closing point is a big duration cam with a big LSA - where the big LSA is often spec'd along with the big duration to avoid too much overlap!
Area Under the Lift Curve: Internet cam experts talk about "maximizing the area under the lift curve" - adding up the valve lift at 1 degree increments to get a sort of "total lift" measure for a cam profile. Its not clear how meaningful this number is, and you can't accurately compute it with just a single lift and duration number - you need to know the exact cam profile, and cam grinders are often pretty secretive about this.
But never underestimate a geek with a spreadsheet... I came up with approximate cam profiles to try to get a general idea of how lift and duration affect the area under the curve. And what I found wasn't very surprising: increasing lift 10% increases the area under the curve about 10%, and increasing duration 10% also increases the area-under-the curve about 10%. A couple things were interesting (and I don't know how meaningful any of this is, as there was lots of approximating going on):
- When I limited the opening ramp rates to what I think are realistic values for a street cam (.005 inches per crankshaft degree), I found that trying to squeeze a lot of lift into too little duration did little to increase the area-under-the-curve. This should have been obvious: a cam with 200 degrees .050 duration and .45 inches of lift takes about 80 crankshaft degrees to go from .050 to (nearly) full lift, and another 80 degrees to close - leaving 40 degrees in the middle of nearly full open. Compare that to a cam with the same duration and .50 inches of lift: because the steepness of the opening and closing ramps are limited the first and last 80 degrees of lift will be about the same, only for the 40 degrees in the middle will the valve be open further than the smaller lift cam and this will only add 2-3% to the area-under-the-curve.
- Even for a fairly long duration cam (240 degrees at .050) most of the area under the curve - 90% - comes during the 180 degrees of the actual intake stroke. For all the debate about the impact of closing the intake late or opening it early, the valve is barely open (very little lift) before TDC or after BDC - it seems that much of the advantage of a large duration cam comes from being able to open the valve wider during the actual intake stroke, when the piston is actually "creating vacuum" in the cylinder. This suggests an asymmetric cam profile to open the valve quickly may be more important than a cam with a super long duration.
Increased lift and near stock duration - e.g. the Isky L-475 grind (0.475 lift and 222 degrees .050 duration), sometimes referred to as a "stage I" cam, with (probably) 109 degree LCAs. This is a pretty mild cam with almost no .050 overlap; the extra lift and small increase in duration will increase flow a bit without radical changes in valve-timing. This cam is a small step up from stock without hurting drivability; it is also a good choice for a street driven engine with individual-runner carbs where a late intake-valve closing point can cause low RPM reversion.
Near stock lift and slightly longer duration - e.g. a Web Camshafts 93a grind with .430 inches of lift and 230 degrees .050 duration, with 11 degrees of .050 overlap. The intake valve closes at 43 degrees ABDC, about 14 degrees more than stock but fairly mild compared to the "hot" street cams. Such a cam has the advantage of working with stock valve springs (but who is going to use 40 year old springs with a new camshaft?). The longer valve duration will increase total air flow even with smallish (stock) carbs and manifolding, and the later intake valve closing point will push the power band up about 1000 RPM. This should perk up a nearly stock L28 with carbs (the overlap may cause problems with the FI manifold) but there isn't enough lift to make really big numbers.
Next we come to the higher lift and longer duration cams - e.g. an Isky L-490 (0.490 lift and 242 degrees .050 duration) or Web 155a grind (0.488 lift and 240 degrees duration). This type of cam is often called a "stage III" cam, and it tries to walk the line between a street cam and an all out race cam; you'll need a better than stock intake and exhaust system to take advantage of the extra lift.
Historical note: back in the 1970s Isky worked with Datsun to produce a line of performance cams for street and racing applications: the Isky Stage I (or L-475) cam was sold by Datsun as the "L7" grind, and was the hot setup on the street - as mentioned it is a fine step up for an otherwise stock engine. The Isky Stage III (or L-490) cam was known as the "L9", and was intended for all out street engines or mild racing applications with ported heads. The Isky L-480 falls somewhere in between these two Datsun labeled grinds with 230 degrees duration, and you might think there was a Datsun "L8" grind corresponding to the L-480 Isky cam, but that didn't happen - the "L8" name was used for some other Datsun performance part.
Finally there are the "mild competition" cams - e.g. the Web "87" and Isky Z-196 grind. These cams have slightly more than .500 inches of lift, slightly more than 240 degrees .050 duration and an LSA around 106 degrees. While the lift and duration are only slightly higher than the "stage III" cams mentioned above, they combine with the narrower lobe-centers to create significantly more overlap - about 35 degrees vs 24 degrees - without changing the intake-valve-close point. These cams would be ideal for an L6 with an individual runner intake (triple side-draft carbs or ITB fuel injection), but have a little too much overlap for a smooth idle in an L6 with dual SUs.
Both Isky and Web (and others) offer even bigger cams - as much as 0.600 lift and 280 degrees .050 duration - but these are generally characterized as race-only camshafts: they have a lot (60+ degrees at .050) of overlap and a very late intake valve close, and while you might get by with some of the milder competition grinds on the street they aren't likely to be much fun to drive.
What is elusive is the middle ground between stage II and III; a cam that makes lots of torque in the low to mid-range, without falling flat on its face past 5500RPM, all with a reasonable idle. Trying to do two things at once usually means not doing either one especially well, but that is the dilemma of choosing a camshaft. There are two approaches to this problem, both a bit of a science project.
Another option is to run a tighter LSA, especially for a carbureted street engine - with either stock SUs or individual runner side-drafts - that doesn't have to pass emissions. The idea is to combine a goodly amount of duration with a slightly tighter LSA to create a bit of overlap (and so punch-up the mid-RPM torque) without changing the intake close point so much as to kill the top end.
Lets say we start with an Isky L-480 (230 degrees at .050) or L-490 (242 degrees) and spec the intake LCA at 104 degrees and the exhaust LCA to 106 for a 105 degree LSA,. This will give either 10 or 20 degrees of .050 overlap, and an intake valve close point of either 34 or 39 degrees ABDC (the incredibly mild stock cam has an intake close around 29 degrees ABDC). This cam should provide a reasonable dynamic compression ratio with a static CR in the mid 9:1 range;
True confession time: I took this path, starting with a Web camshaft 155a grind (similar to the Isky L-490) and having them cut it with a 106 degree LSA. The engine is still being broken in, and hasn't been fine-tuned or dyno'd yet, but it does indeed pull hard from 3000-6000RPM (and wants to keep going). But sadly, the cold idle is pretty rough . It may be possible to tune a slightly better idle, but I'm starting to regret trying to out-engineer Isky and not just pick their L-6 grind.
Hope this has helped shed some light - your opinions and feedback are welcome!
The Isky L-6 is a bit of an outlier; despite the name it doesn't seem to have been offered in the Datsun catalog. The L-6 has .540 inches of lift and 230 degrees duration on 109 lobe centers. The high lift works well with engines that have had the head ported and have the carburetion and exhaust to support high RPM. While the duration is somewhat modest, it is an asymmetric cam that compensates a bit, without creating a huge amount of overlap or delaying the intake close point, making for a more street friendly package - think of it as a stage 2.5 grind. By all accounts it is a good cam for a maximum effort street L28 - you just have to be careful of valve to piston clearance and having proper springs (available from Isky) for the relatively high-lift.
Another option is to run a tighter LSA, especially for a carbureted street engine - with either stock SUs or individual runner side-drafts - that doesn't have to pass emissions. The idea is to combine a goodly amount of duration with a slightly tighter LSA to create a bit of overlap (and so punch-up the mid-RPM torque) without changing the intake close point so much as to kill the top end.
Lets say we start with an Isky L-480 (230 degrees at .050) or L-490 (242 degrees) and spec the intake LCA at 104 degrees and the exhaust LCA to 106 for a 105 degree LSA,. This will give either 10 or 20 degrees of .050 overlap, and an intake valve close point of either 34 or 39 degrees ABDC (the incredibly mild stock cam has an intake close around 29 degrees ABDC). This cam should provide a reasonable dynamic compression ratio with a static CR in the mid 9:1 range;
True confession time: I took this path, starting with a Web camshaft 155a grind (similar to the Isky L-490) and having them cut it with a 106 degree LSA. The engine is still being broken in, and hasn't been fine-tuned or dyno'd yet, but it does indeed pull hard from 3000-6000RPM (and wants to keep going). But sadly, the cold idle is pretty rough . It may be possible to tune a slightly better idle, but I'm starting to regret trying to out-engineer Isky and not just pick their L-6 grind.
CWC cam blanks 😱 When you buy a camshaft, you'll usually have a choice between:
- Buying a new cam "blank" to have the specific profile you've chosen ground onto.
- Regrinding a factory Nissan camshaft.
A regrind is cheaper (you're not paying for the cam blank) but involves compromises. It is difficult (not impossible - but expensive) to add metal to an existing cam lobe. More typically, if you want to increase the lift of a cam you make the "base circle" - the circular half of the lobe opposite the nose that doesn't quite touch the cam follower - smaller than original. That allows the rockers to be adjusted up - closer to the center of the camshaft, so that the rocker moves further when the stock radius "nose" is centered on the rocker contact surface. Increasing the duration means flattening the nose of the stock lobe, which by itself reduces lift and requires taking still more metal off the base circle. Changing the LSA or creating a more asymmetric profile means shifting the nose of the lobe, again reducing lift and requiring a smaller base circle. Of course there is a limit on how small the base circle can be, that limits how much the original profile can be changed, but it is common to grind a cam where the base circle is slightly smaller (a few hundredths of an inch) than the shaft itself.
Reducing the size of the base-circle is often considered undesirable, although why isn't clear. Most likely it makes the transitions between the opening ramp and full open parts of the valve-lift profile less smooth - that can lead to oscillations in the spring at high RPM, and probably doesn't help the life of the springs and camshaft. It also changes the geometry of the rocker arms with respect to the valve, typically requiring taller lash-caps to keep the lobe contact on the rocker contact patch. While the practice might be less than ideal, it is widely done.
The alternative is to start with a new "blank" - the rough casting that is machined to the desired profile. Blanks start out with grossly oversize lobes that allow for a great deal of variation in lift, duration, LSA and profile while maintaining as large a base circle as possible. In the past, cam blanks were widely available and modestly priced, so using them was the preferred approach, especially for a high performance grind with large lift and duration.
Unfortunately, the Nissan L6 has been out of production for 40ish years, and the supply of cam blanks has dried up. OEM blanks were originally available from Nissan, and are considered the best quality, but they are practically non-existent. In the past other manufacturers have produced blanks for the L6 to feed the demand for replacement parts, most notably Estas (located in Turkey) and CWC (Camble, Wyant and Canton in the US, a division of Kautex that is owned by the Textron conglomerate). It seems neither of these companies are still producing L6 blanks, whatever is out there is all that are left.
Estas makes camshafts for many European auto makers, their blanks are "chilled iron" and have been used by some of the reputable L6 engine builders. CWC is a major supplier of camshafts to US auto makers, their blanks are sand cast and induction hardened.
While CWC blanks have historically been used successfully in the Nissan L6, in recent years they have earned a reputation for poor quality - where some (but generally not all) of the lobes experience rapid wear - in just a few 1000 miles of operation - especially when used in high performance applications. The speculation is that the last run of CWC blanks were not properly hardened. Its not clear if all CWC blanks suffer this problem - or if just some small fraction slipped through QA and of course the failures are the ones you hear about. Some of the cam grinders say they give the CWC blanks an extra heat treatment before grinding them to correct improper hardening.
The future of the performance cam industry is likely to be CNC machined steel parts, but that technology hasn't (yet) filtered down to the limited market for L6 engines. Until then, having the original Nissan cam reground has the least risk.
Estas makes camshafts for many European auto makers, their blanks are "chilled iron" and have been used by some of the reputable L6 engine builders. CWC is a major supplier of camshafts to US auto makers, their blanks are sand cast and induction hardened.
While CWC blanks have historically been used successfully in the Nissan L6, in recent years they have earned a reputation for poor quality - where some (but generally not all) of the lobes experience rapid wear - in just a few 1000 miles of operation - especially when used in high performance applications. The speculation is that the last run of CWC blanks were not properly hardened. Its not clear if all CWC blanks suffer this problem - or if just some small fraction slipped through QA and of course the failures are the ones you hear about. Some of the cam grinders say they give the CWC blanks an extra heat treatment before grinding them to correct improper hardening.
The future of the performance cam industry is likely to be CNC machined steel parts, but that technology hasn't (yet) filtered down to the limited market for L6 engines. Until then, having the original Nissan cam reground has the least risk.
Hope this has helped shed some light - your opinions and feedback are welcome!
About Compression Ratios
It is well known that when it comes to automobile engines more compression makes more power, but the reasons why - at least on the internet - tend to be a little handy-wavy. The common explanation is that higher compression makes higher pressure in the combustion chamber, which means the burning air-fuel mix pushes harder on the piston. While that is somewhat true, it ignores the fact that compressing the air-fuel mix more also steals more energy during the compression stroke, and that cancels out some of the extra power produced due to higher pressure on the power stroke.
Saturday, April 3, 2021
Lies, Damn Lies and Horsepower
Anyone who has ever set out to plan an engine build has had to deal with the mysteries of horsepower and torque. The internet is full of information - some good, some not so much - trying to explain what these things are and why they are important. Before I start talking about engine mods I'm going to try just this once to nail down the science.
Sunday, August 16, 2020
How Fast is Fast?
There is an old saying among hot-rodders and racers:
"Speed costs money. How fast can you afford to go?"
Honestly answering that question is the first step in deciding what sort of car and engine you're setting out to build, else you will likely spiral into a constant game of second guessing yourself - looking for more and more power - that at best will cause you to spend more than you have to and at worst will prevent you from ever actually finishing your car!
It takes at most 50 horsepower for a typical mid-size car to cruise at 100 mph. The BMW Z3 coupe - with similar size and aerodynamics as our beloved S30, and with 228 crank horsepower (in the ballpark of a stout L28 street build) - has a top speed of 155mph. Any modern car can easily reach the century mark - even though in most of the US you can't drive faster than about 85 mph without losing your license. On the street, top speed is not nearly as important as acceleration: how quickly can you speed up to change lanes or merge into highway traffic or pass a lumbering truck on a two lane road.
Note: this article is aimed at street driven cars; there may be some useful info for road racers and drag racers, but if that's where your interest lies you'll have to fill in some gaps on your own.
Lets start with the basics. If you drop a rock off a tall building gravity pulls it towards the ground causing the rock to accelerate - increase it's speed - at a constant rate: every second the rock falls its downward speed increases by 32 feet-per-second until it actually hits the ground. After 1 second the rock is falling at a speed of 32 feet/sec, after 2 seconds its falling at 64 feet/sec, and after 2.75 seconds it is falling at 88 feet/sec, which happens to be 60mph. We call the acceleration of gravity - 32 feet-per-second-per-second - 1 gravity - or 1G (its really 32.2 ft/s^2, but 32 is close enough for government work).
When it comes to cars, the force applied by the drive wheels will cause the car to accelerate forward.
Automotive acceleration is traditionally measured by 0-60mph times (or 0-100kph times if you prefer metric units). Flat-out 0-60mph sprints aren't very common in every day driving; it was just an easy to measure number that the car magazines latched onto and has taken on an outsized importance. Still, it gives us a place to start.
It takes at most 50 horsepower for a typical mid-size car to cruise at 100 mph. The BMW Z3 coupe - with similar size and aerodynamics as our beloved S30, and with 228 crank horsepower (in the ballpark of a stout L28 street build) - has a top speed of 155mph. Any modern car can easily reach the century mark - even though in most of the US you can't drive faster than about 85 mph without losing your license. On the street, top speed is not nearly as important as acceleration: how quickly can you speed up to change lanes or merge into highway traffic or pass a lumbering truck on a two lane road.
Note: this article is aimed at street driven cars; there may be some useful info for road racers and drag racers, but if that's where your interest lies you'll have to fill in some gaps on your own.
Lets start with the basics. If you drop a rock off a tall building gravity pulls it towards the ground causing the rock to accelerate - increase it's speed - at a constant rate: every second the rock falls its downward speed increases by 32 feet-per-second until it actually hits the ground. After 1 second the rock is falling at a speed of 32 feet/sec, after 2 seconds its falling at 64 feet/sec, and after 2.75 seconds it is falling at 88 feet/sec, which happens to be 60mph. We call the acceleration of gravity - 32 feet-per-second-per-second - 1 gravity - or 1G (its really 32.2 ft/s^2, but 32 is close enough for government work).
When it comes to cars, the force applied by the drive wheels will cause the car to accelerate forward.
Automotive acceleration is traditionally measured by 0-60mph times (or 0-100kph times if you prefer metric units). Flat-out 0-60mph sprints aren't very common in every day driving; it was just an easy to measure number that the car magazines latched onto and has taken on an outsized importance. Still, it gives us a place to start.
If we assume acceleration is constant, its easy enough to convert between 0-60mph times and acceleration in physical units such as feet/sec^2 or G's:
acceleration = change-in-speed / time, or re-arrange: time = change-in-speed / acceleration
Units are always important: 60mph is 88 feet/second, and G is 32 feet/second/second. A real world example: Car&Driver got the (somewhat porky - 2900 pound) 2017 Civic Si to 60mph in 6.7 seconds:
acceleration = 88 ft/sec / 6.7 seconds = 13.1 feet/sec/sec, 13.1 / 32 = .41G
Thanks to modern technology even the most under-powered econobox (I'm looking at you Toyota Yaris) can manage 0.3G of acceleration.
0-60 time = 88 feet/second / (0.3 x 32 feet/second/second) = 9.2 seconds
Plugging in a few different numbers for acceleration gives us a chart of equivalent 0-60 times:
The fastest of today's FWD hot-hatches can do 0-60 in the low 6 second range, the BMW Z3 I mentioned earlier manages it in 6-seconds flat, modern muscle cars like a Mustang or Camaro can do it in a hair under 4 seconds, and some of the quickest super-cars can do 0-60 in the low 2s - they can accelerate forward faster than a falling rock! By comparison, 1970s Porsche 911s went 0-60 in what is today a pretty pedestrian 7.5 seconds, and when showroom new our Z cars were in the mid 8-second range!
Acceleration is limited by both tire traction and power. Lets take those one at a time.
Traction comes down to friction between the tire and the road. A lot of college physics books claim the coefficient of friction for rubber on concrete is 0.8, meaning a rubber tire can apply a maximum forward force of 0.8 times the weight pressing down on that tire before it starts to slip/spin. If you do a little physics, for a car with 2 driven wheels carrying 50% of the weight of the car, the maximum acceleration in G's is half the coefficient of friction - or in this example about 0.4G - equivalent to a 6.9 second 0-60 time.
If that is the traction limit, how does any car ever go 0-60 in less than 6.9 seconds? The model of friction in those physics books is a bit of an oversimplification based on one hard smooth surface resting on another. A 0.8 coefficient of friction is realistic for the hard narrow tires that were state of the art in the 1960s and 70s when a lot of those physics books were written - but wider tires and softer rubber can yield higher coefficients - even greater than 1.0 (something some introductory physics books suggest is impossible). Modern performance-oriented street tires have coefficients in the ballpark of 1.0-1.1 and extra wide ultra-soft compound "max-performance" tires can push that to at least 1.3. The 2018 V8 powered Ford Mustangs have 12 inch wide performance tires on the back and manage .7G of acceleration.
Another piece of this puzzle is weight distribution. Putting more of the car's weight on the driving wheels means those wheels can deliver more forward force on the car, yielding greater acceleration. Mid-engine supercars put about 60% of their weight on the rear driving wheels, which combined with super-wide super-soft tires is a big part of how they achieve sub-3 second 0-60 times. Our Z's have close to a 50/50 front/rear weight distribution, thanks mostly to the big gas-tank in the rear, which is pretty good for a front-engine rear-wheel-drive car. With the engine and transaxle up front, a front-wheel-drive car puts 60% (or more) of its weight on the driving wheels, compared to 50% or less for a front-engine RWD car, giving FWD an edge. Under acceleration, weight is transferred to the rear wheels, which helps a RWD car and hurts the FWD car, but for practical purposes the transfer isn't enough to offset the FWD advantage.
Having 4 driven wheels is better than 2, allowing 1G or more of acceleration on street tires. This is how the Subaru STIs can rival the Mustang's 0-60 times without the big V8 horsepower (the Subie does 0-60 in 4.6 seconds with "just" 310 crank horsepower); unfortunately 4WD is not exactly a bolt-on mod.
Now lets talk about power. Acceleration requires power, and the power required goes up with both the car's speed and acceleration. At low speeds where air drag is negligible there is a simple relationship:
Power = mass x velocity x acceleration
We can use this to get a pretty good estimate of the power required for a 0-60mph dash. Its easiest to use metric units:
Power = 1200Kg x 26.8m/sec x (.47 x 9.8m/sec/sec) = 148,129 Watts
Divide by 745.7 Watts per horsepower to get 199 (wheel) horsepower.
This is the power needed to accelerate (at .47G) the last little bit to 60mph; it is the maximum horsepower needed in a .47G 0-60mph sprint. This assumes the car is geared to deliver that 199hp at 60mph, and that the driver is able to keep the throttle right on the edge of wheel spin the whole way to 60. I'll dig into the gearing part of this story in another post but for now we have a way to estimate what we need in the way of tires and engine mods to reach a particular 0-60 target.
So you might be thinking "a 6-second 0-60 time is pretty darn quick, why does anyone build an s30 with more than 200whp?" Which reminds me of an old joke about dogs with the punchline "Because they can!" Having more power, especially in the mid-RPMs (what we usually mean when we think of a torquey engine) can give us more acceleration in 3rd or 4th gear at speeds beyond 60mph - e.g. for making a quick pass on the interstate. But mostly, you only need more power if you are racing and need lots of acceleration beyond 100mph.
Or if you have a lot of traction. A new V8 Mustang with all the go fast options can do 0-60mph in about 3.9 seconds (.7G of acceleration). Not surprisingly, the Mustang is packing 460 crank horsepower and 12inch wide rear tires - about 50% wider than the typical Z-car tire. A Z-car with similar 12 inch tires (and fender flares to clear them) should have enough traction to match the Mustang, but lets look at the power requirement:
Power = 1200Kg x 26.8m/sec x (.7 x 9.8m/sec/sec) = 220,618 Watts = 296whp
Assuming a 15% power loss in drive-train friction, that's about 350 crank horsepower - a lot to squeeze from a streetable 2.8 liter NA engine.
Real world complications: When you dig into the numbers for a Z car with typical NA power levels, you find that the cars are usually traction limited in 1st gear and power limited in 2nd gear. If you take the car with 200whp from the example above and fit a wider rear tire - lets say an "ultra performance" 225mm tire - the car will be able to accelerate harder in 1st gear and turn a quicker 0-60 without adding more horsepower. The math is more complicated and depends on gearing and how much stickier the wider tire is, but increasing max acceleration from .47 to .55G while keeping the same 200whp can knock about 0.4 seconds off the 6 second 0-60 time. The thing is this is mostly 1st gear improvement: without more horsepower the bigger tires won't generate any more acceleration in 2nd gear or above; the wider tires help at the drag strip but won't help with merging at the end of the on-ramp.
Back to How Fast is Fast? What we usually think of as "speedy" in a street car - after "can it spin the tires" - is how well the car accelerates when downshifting a gear for a quick pass. Without a turbo or big V8 displacement, beyond 2nd gear and 60mph there just isn't enough power available from an NA L6 for extreme acceleration, especially given the rapidly increasing aerodynamic drag soaking up more and more power. In general the more horsepower the engine can provide the better - again with the caveat that the gearing needs to allow the engine to make the power where its needed.
Lets look at this scenario in a little more detail: if we had that hypothetical 200whp L-6, what acceleration do we see in 4th gear at 75mph? Gearing plays a part here, but 4th gear at 70mph corresponds to about 4000RPM, and about 120whp.
The big take-aways from all of that is that trying to build an NA L6 Z-car that can keep up with a modern performance car is hard; it requires an impractical amount of horsepower from the L6, really wide tires, and most of that performance can't really be used on the street. If that kind of performance is your goal you really need a turbo engine or V8 swap. But making enough horsepower to be quicker than 99% of everyday traffic isn't all that hard, and can make for a fun-to-drive car - and that 200whp is what I am aiming for.
There are also a few surprises when we look at gearing, but that's another story.
acceleration = change-in-speed / time, or re-arrange: time = change-in-speed / acceleration
Units are always important: 60mph is 88 feet/second, and G is 32 feet/second/second. A real world example: Car&Driver got the (somewhat porky - 2900 pound) 2017 Civic Si to 60mph in 6.7 seconds:
acceleration = 88 ft/sec / 6.7 seconds = 13.1 feet/sec/sec, 13.1 / 32 = .41G
Thanks to modern technology even the most under-powered econobox (I'm looking at you Toyota Yaris) can manage 0.3G of acceleration.
0-60 time = 88 feet/second / (0.3 x 32 feet/second/second) = 9.2 seconds
Plugging in a few different numbers for acceleration gives us a chart of equivalent 0-60 times:
- 0.3G = 9.2 seconds
- 0.4G = 6.9 seconds
- 0.47G = 6 seconds
- 0.5G = 5.5 seconds
- 0.55G = 5 seconds
- 0.6G = 4.6 seconds
- 0.7G = 3.9 seconds
The fastest of today's FWD hot-hatches can do 0-60 in the low 6 second range, the BMW Z3 I mentioned earlier manages it in 6-seconds flat, modern muscle cars like a Mustang or Camaro can do it in a hair under 4 seconds, and some of the quickest super-cars can do 0-60 in the low 2s - they can accelerate forward faster than a falling rock! By comparison, 1970s Porsche 911s went 0-60 in what is today a pretty pedestrian 7.5 seconds, and when showroom new our Z cars were in the mid 8-second range!
Acceleration is limited by both tire traction and power. Lets take those one at a time.
Traction comes down to friction between the tire and the road. A lot of college physics books claim the coefficient of friction for rubber on concrete is 0.8, meaning a rubber tire can apply a maximum forward force of 0.8 times the weight pressing down on that tire before it starts to slip/spin. If you do a little physics, for a car with 2 driven wheels carrying 50% of the weight of the car, the maximum acceleration in G's is half the coefficient of friction - or in this example about 0.4G - equivalent to a 6.9 second 0-60 time.
If that is the traction limit, how does any car ever go 0-60 in less than 6.9 seconds? The model of friction in those physics books is a bit of an oversimplification based on one hard smooth surface resting on another. A 0.8 coefficient of friction is realistic for the hard narrow tires that were state of the art in the 1960s and 70s when a lot of those physics books were written - but wider tires and softer rubber can yield higher coefficients - even greater than 1.0 (something some introductory physics books suggest is impossible). Modern performance-oriented street tires have coefficients in the ballpark of 1.0-1.1 and extra wide ultra-soft compound "max-performance" tires can push that to at least 1.3. The 2018 V8 powered Ford Mustangs have 12 inch wide performance tires on the back and manage .7G of acceleration.
Another piece of this puzzle is weight distribution. Putting more of the car's weight on the driving wheels means those wheels can deliver more forward force on the car, yielding greater acceleration. Mid-engine supercars put about 60% of their weight on the rear driving wheels, which combined with super-wide super-soft tires is a big part of how they achieve sub-3 second 0-60 times. Our Z's have close to a 50/50 front/rear weight distribution, thanks mostly to the big gas-tank in the rear, which is pretty good for a front-engine rear-wheel-drive car. With the engine and transaxle up front, a front-wheel-drive car puts 60% (or more) of its weight on the driving wheels, compared to 50% or less for a front-engine RWD car, giving FWD an edge. Under acceleration, weight is transferred to the rear wheels, which helps a RWD car and hurts the FWD car, but for practical purposes the transfer isn't enough to offset the FWD advantage.
Having 4 driven wheels is better than 2, allowing 1G or more of acceleration on street tires. This is how the Subaru STIs can rival the Mustang's 0-60 times without the big V8 horsepower (the Subie does 0-60 in 4.6 seconds with "just" 310 crank horsepower); unfortunately 4WD is not exactly a bolt-on mod.
Now lets talk about power. Acceleration requires power, and the power required goes up with both the car's speed and acceleration. At low speeds where air drag is negligible there is a simple relationship:
Power = mass x velocity x acceleration
We can use this to get a pretty good estimate of the power required for a 0-60mph dash. Its easiest to use metric units:
- A 2600 pound Z-car masses 1200Kg.
- 60mph is 26.8 meters/second
- 1G is 9.8 meter/second/second
- 1 horsepower is 745.7 Watts
Power = 1200Kg x 26.8m/sec x (.47 x 9.8m/sec/sec) = 148,129 Watts
Divide by 745.7 Watts per horsepower to get 199 (wheel) horsepower.
This is the power needed to accelerate (at .47G) the last little bit to 60mph; it is the maximum horsepower needed in a .47G 0-60mph sprint. This assumes the car is geared to deliver that 199hp at 60mph, and that the driver is able to keep the throttle right on the edge of wheel spin the whole way to 60. I'll dig into the gearing part of this story in another post but for now we have a way to estimate what we need in the way of tires and engine mods to reach a particular 0-60 target.
So you might be thinking "a 6-second 0-60 time is pretty darn quick, why does anyone build an s30 with more than 200whp?" Which reminds me of an old joke about dogs with the punchline "Because they can!" Having more power, especially in the mid-RPMs (what we usually mean when we think of a torquey engine) can give us more acceleration in 3rd or 4th gear at speeds beyond 60mph - e.g. for making a quick pass on the interstate. But mostly, you only need more power if you are racing and need lots of acceleration beyond 100mph.
Or if you have a lot of traction. A new V8 Mustang with all the go fast options can do 0-60mph in about 3.9 seconds (.7G of acceleration). Not surprisingly, the Mustang is packing 460 crank horsepower and 12inch wide rear tires - about 50% wider than the typical Z-car tire. A Z-car with similar 12 inch tires (and fender flares to clear them) should have enough traction to match the Mustang, but lets look at the power requirement:
Power = 1200Kg x 26.8m/sec x (.7 x 9.8m/sec/sec) = 220,618 Watts = 296whp
Assuming a 15% power loss in drive-train friction, that's about 350 crank horsepower - a lot to squeeze from a streetable 2.8 liter NA engine.
Real world complications: When you dig into the numbers for a Z car with typical NA power levels, you find that the cars are usually traction limited in 1st gear and power limited in 2nd gear. If you take the car with 200whp from the example above and fit a wider rear tire - lets say an "ultra performance" 225mm tire - the car will be able to accelerate harder in 1st gear and turn a quicker 0-60 without adding more horsepower. The math is more complicated and depends on gearing and how much stickier the wider tire is, but increasing max acceleration from .47 to .55G while keeping the same 200whp can knock about 0.4 seconds off the 6 second 0-60 time. The thing is this is mostly 1st gear improvement: without more horsepower the bigger tires won't generate any more acceleration in 2nd gear or above; the wider tires help at the drag strip but won't help with merging at the end of the on-ramp.
Back to How Fast is Fast? What we usually think of as "speedy" in a street car - after "can it spin the tires" - is how well the car accelerates when downshifting a gear for a quick pass. Without a turbo or big V8 displacement, beyond 2nd gear and 60mph there just isn't enough power available from an NA L6 for extreme acceleration, especially given the rapidly increasing aerodynamic drag soaking up more and more power. In general the more horsepower the engine can provide the better - again with the caveat that the gearing needs to allow the engine to make the power where its needed.
Lets look at this scenario in a little more detail: if we had that hypothetical 200whp L-6, what acceleration do we see in 4th gear at 75mph? Gearing plays a part here, but 4th gear at 70mph corresponds to about 4000RPM, and about 120whp.
- 120hp = 89500 watts
- mass of the car is stil 1200Kg
- 75mph = 33.5 meters/sec
Rearrange the formula:
acc = power / (mass * velocity) = 89500 W / (1200Kg * 33.5 m/s) = 2.2 meters/sec/sec = .22G
This doesn't sound like a lot, but remember, its about twice what typical traffic reaches. Having enough power to spin the tires at highway speeds sounds cool, but there is no good reason to do it and its a good way to lose control of the car. Having a bit more acceleration might be nice - and if you really need it, it is available in 3rd gear (75mph in 3rd is about 5200RPM, near a typical L6 power peak).
acc = power / (mass * velocity) = 89500 W / (1200Kg * 33.5 m/s) = 2.2 meters/sec/sec = .22G
This doesn't sound like a lot, but remember, its about twice what typical traffic reaches. Having enough power to spin the tires at highway speeds sounds cool, but there is no good reason to do it and its a good way to lose control of the car. Having a bit more acceleration might be nice - and if you really need it, it is available in 3rd gear (75mph in 3rd is about 5200RPM, near a typical L6 power peak).
The big take-aways from all of that is that trying to build an NA L6 Z-car that can keep up with a modern performance car is hard; it requires an impractical amount of horsepower from the L6, really wide tires, and most of that performance can't really be used on the street. If that kind of performance is your goal you really need a turbo engine or V8 swap. But making enough horsepower to be quicker than 99% of everyday traffic isn't all that hard, and can make for a fun-to-drive car - and that 200whp is what I am aiming for.
There are also a few surprises when we look at gearing, but that's another story.
Wednesday, July 29, 2020
Do you really want a silk purse?
When I was a snot-nosed kid growing up in small town Pennsylvania, having your own car was the Holy Grail of high school life. Unless your parents were especially well-off and especially generous, having a car at age 17 meant saving up several years of less than minimum wage income, buying a worn-out car and then learning to wrench it back to life. One of the things we learned kind of quickly was that the actual wrenches came in 3 varieties:
- No-name stuff they sold at the no-name discount store that were really cheap and quickly broke.
- The shiny Proto tools hanging on a big display behind the counter at the NAPA store; the chrome would have looked at home on a '57 Chevy. These were really pretty and never broke, but you could spend a whole week's paper-route money on a single socket!
- And then there were Craftsman brand tools from the Sears store. Less than half the price of the Proto stuff, not quite as shiny but with the same lifetime guarantee (this was back in 1976 - sadly things have changed - the quality of Craftsman tools has slipped and Sears itself is on shaky ground). Like a lot of my friends, the Christmas after I turned 16 I got the 100-piece Craftsman tool set under the tree - 1/4, 3/8 and 1/2 inch drive ratchets and sockets and a set of combination wrenches in a nice steel toolbox. I think back in 1976 that set cost all of $99. And I still have most of those wrenches today!
What does this little trip down memory lane have to do with old Datsuns? Simple enough, whatever part you want to buy or repair you want to make, there will usually be 3 options: something cheap that will quickly wear out or break, something fancy and really expensive, and something that is not cheap or fancy but will work well and last forever.
A lot of the folks you'll find working on old cars have technical backgrounds: there are the unsurprising machinists and mechanics and engineers, but also computer programmers and finance guys and various kinds of health care workers. They tend to be the people who have to be fanatical about doing thieir job just right. Doing a half-way job on a car that you are passionate about is unthinkable, but doing everything to perfection uaully means never getting to the point of having a car that you can drive and have fun showing off and taking places. When you're planning out a restoration or engine build it is tempting to go for all the fancy stuff. For any one part its often just a few dollars more, and when you're going back and forth between which thing to buy you start thinking "I'm only going to buy it once, I might as well go for the best". The hidden gotcha is that by adding 20% (or more) to the cost of everything, you're likely stretching out the day when you have the money to actually finish the car - if you don't lose interest before that happens!Its not just money, sometimes its actually time. This occured to me in the middle of a project to spiff up the interior of my 260Z. I had bought new seat covers and foam seat inserts from Banzai Motorworks and the fellow who runs the place recommended ACC Carpets. They had reasonable prices, so I bought a set - no hassles, came quickly. Then I pulled the old carpets out, and while the floors were still pretty solid there was some ugly rust where the floor pans were welded together and a spot under the passenger seat where I suspect someone spilled something corrosive (maybe a Coke?) many years ago. No problem, I think - I'll just use one of the rust converters to clean them up and put a coat of "rust encapsulating" paint on them before I put the carpets in.
After a couple weekends this job just got bigger and bigger. The jute padding on the transmission tunnel smelled like a wet dog - I pulled it up and used Goof-Off glue remover to cleanup the paint underneath. The sound deadening "tar mats" were removed - through a combination of dry ice (about $50 worth), heatgun (had to buy that too) and a putty knife (I thankfully already had). I read up on rust-converters and bought a gallon jug of Evaporust and followed the directions: spread paper towels over the rusty floors, soak it with the Evaporust liquid and cover with plastic garbage bags to keep it from drying out. The rust came up, but slowly. Sanded some of the worst pitted areas and applied a heavier dose of Evaporust. Started to see bare metal. Switched to MetalBlast, a phosphoric acid based product that removes and neutralizes rust and etches the steel in preparation for paint (it also removed a lot of the remaining factory paint). And finally I brushed on two coats of RustBullet.
 |
| Before |
 |
| Evaporust followed by Metal Blast |
 |
| A lot of rust removed |
 |
| Two coats of Rust Bullet |
When I bought my Z, I searched hard to find a fairly complete car with a solid body and minimal wear and tear. I set an informal search radius of 200 miles, and spent a lot of weekends driving to look at Datsuns. Along the way I passed on a few nice looking and nice driving cars that were either patched-up (wheel openings made largely of Bondo) or where the seller was obviously misrepresenting the condition of the car (e.g. claiming original paint on an obvious re-spray). Eventually I found a 260Z with not too much rust in the usual places, with a nice looking but not too-recent re-spray and an interior you weren't afraid to get into without a recent tetanus shot. Sure, the engine ran a bit rough, the AC didn't work and the suspension had a metal-on-metal feel, but all the pieces were there. The repairs it needed were things that bolted on.
If you read the collector car buyer guides, I did all the right things, but looking back I did something very wrong: I bought a car that you could drive but that no one really wanted to drive. Getting the car to that driver level - not just moving under its own power with the road manners of a worn out pickup truck, but feeling like an honest-to-God sports car - was a lot of work and money, where all of those there-but-nearly-worn-out parts need to be replaced anyhow. And if you don't have the money, time and facilities to get to that driver level, you are likely to end up with a neglected project car moldering in your garage.
It's too late for me, but if you're just starting down the collector car path, do yourself a favor: either buy a car that truly drives well, or if you can't get the perfection-monkey off your back start with a basket case car where you know you will have to completely replace everything. Trying to walk the middle of the road is a good way to get run over!
Thursday, July 9, 2020
One day at a time
In the software engineering world there is a famous book - The Mythical Man Month by Fred Brooks - that tells of the trials and tribulations of one of the first big software projects (the development of the IBM 360 operating system). The book has a joke that has become famous in the software biz:
Q: "How does a project get to be a year late?"
A: "One day at a time!"
If you're the least bit interested in engineering you should track down a copy of the book and give it a read, but the point of this joke is that there are an endless list of minor set-backs that delay every project - a key employee has the flu, a part is delayed in shipping, a manager asks for a demonstration, the system fails in some unexpected way and requires rework - and all those tiny, impossible to predict delays simply add up.
Thats how my Datsun 260Z managed to sit unmoving in my garage for most of the last 15 years, often trapped by boxes of household junk. Every now and then I'd get it started and drive it to work or to a car show or something, but it was always a bit of an ordeal. The car didn't run well or drive well and it rattled and smelled bad inside.
Q: "How does a project get to be a year late?"
A: "One day at a time!"
If you're the least bit interested in engineering you should track down a copy of the book and give it a read, but the point of this joke is that there are an endless list of minor set-backs that delay every project - a key employee has the flu, a part is delayed in shipping, a manager asks for a demonstration, the system fails in some unexpected way and requires rework - and all those tiny, impossible to predict delays simply add up.
Thats how my Datsun 260Z managed to sit unmoving in my garage for most of the last 15 years, often trapped by boxes of household junk. Every now and then I'd get it started and drive it to work or to a car show or something, but it was always a bit of an ordeal. The car didn't run well or drive well and it rattled and smelled bad inside.
Subscribe to:
Posts (Atom)