Ever since we had the opportunity to drive Endless Racing's Z Car Challenge 400hp naturally aspirated (NA) Z in Japan, building the same engine here has topped our list of geek obsessions. The only reason we never went ahead with it was because Project Z had already been committed for turbocharged power and the cost of building such a fully prepped engine gets pretty outrageous, not to mention an engine like this only drinks 100 octane. But automotive karma works in strange ways.
Castrol invited us to face off against our sister magazines (Super Street, Modified, Eurotuner, Import Tuner, Turbo, Lowrider) by teaming up with a shop of our choice to build a no compromise engine in its Castrol Syntec Top Shop challenge. The competition would be based on peak horsepower and torque per displacement (displacement is multiplied by 2 for anything with forced induction), total horsepower under the curve, a 30 minute endurance death match (all tests will be administered on an engine dyno), and a engineering challenge to impress a panel of engine guru judges. All of which sounds suspiciously like our USCC rules.
While everyone else might be building a dyno queen engine, we're looking to build something real world that might one day end up in some lucky raffle winning schmuck's engine bay (the winning engine will be given away to one lucky member of the public).
Picking a naturally aspirated engine to go into a horsepower battle with turbo engines might not sound like the smartest thing to do, until you look at our rules and consider real world performance. For example, if we chose to build a 3.0-liter Supra motor targeted at 1000 bhp, the numbers would break down to around 166bhp/liter since a turbo motor has it's displacement multiplied by two. A 5.7-liter small block would really suffer even at a wishful 1500bhp. That would land it somewhere around 131bhp/liter, just a little more than Honda's out of the box S2000 engine.
Even if we can't make the most power per displacement with a NA VQ, there's another side of the equation, power delivery or area under the curve. If you've ever seen the power curve of a big turbo, small displacement 1000 hp car, it's essentially useless until the last 1000 rpm and undriveable, regardless of the engine speed. For most of the power band, the engine will be struggling to make a fraction of its peak power until the monster turbo, needed to flow this amount of air, finally spools and skyrockets the power at an uncontrollable rate. Do the math and the area under the power spike won't compare to the steady power coming from a NA engine.
The Engine and the shop
So we decided on Nissan's VQ35DE as our base engine platform. It's the ideal combination of displacement, rpm, fundamental design and flow capabilities, not to mention a respectable amount of low end torque. Add on the 100 octane gas that everyone will be using for the competition and we'll be able to raise the stock compression ratio to a respectable race engine standard. Our aim is to build an easily replicated, street usable, 400bhp naturally aspirated VQ35DE. Several respectable tuners like Tomei, Nismo, Cosworth and Jun Auto already have extensive research and racing programs based around the VQ.
When it came down to picking a shop among the notable VQ tuners, we decided on Cosworth Engineering for several reasons. Namely, they spoke English (even though with a strange accent at times), the US headquarters was right in our back yard here in Southern California, most of the parts we would be using in the build are sold off-the-shelf and, mostly, because few organizations have the engineering capabilities, expertise, and experience to rival Cosworth.
Picking a manufacturer like Cosworth means that we suddenly have a wealth of engineering resources available such as engine simulation software, flow bench data and computational fluid dynamics (CFD) head flow analysis. These guys are real engineers and possess the same tools OEMs use to design an engine from the ground up. This way, in theory, we could run engine simulations of how different bore, stroke, rod length, piston, compression ratio, and displacement combinations would be offer optimal power, torque, flow and engine speeds.
Our hope was to be able to reverse engineer the VQ and analyze the engine from the bottom up to realize why Nissan made certain choices in its design and what could be optimized without sacrificing wear, durability or the innate character of the VQ. At least that's what we hoped. And while Cosworth does have the technology to completely design and manufacture motors from a clean sheet, the labor involved would be beyond the scope of this project. After all this isn't an IRL or F1 engine we're talking about here.
As powerful as all these analysis and modeling tools are, they are still only a guide to be used in conjunction with practical and real world knowledge. So we went with a much more conventional method of design and tuning, and worked around the basic architecture of what Nissan gave us in the VQ.
Real World Tuning
Our high output VQ will be based roughly on Cosworth's concept for a drop in crate motor, which will feature an assortment of parts that have already been released or are already in testing for the VQ. The Top Shop motor will just take it up another notch.
Obviously, since we're using a stock block, our engine design would be constrained by some basic physical limitations. Since the Castrol Syntec Top Shop Challenge makes no restrictions on displacement or flow, we would want to start from the bottom end and maximize our total displacement to take advantage of the lack of a displacement modifier for NA engines.
A simple approach to increasing displacement is to punch out a motor as far as its bore spacing will allow and add as much stroke as possible. This works for old rev-limited V8s with cast iron blocks (where piston speeds aren't such a concern, since reciprocation mass ultimately limits revs). But in a VQ with the potential to spin up to 9000rpm, that's not the case. From a basic displacement perspective, the available room allows Cosworth to increase stock stroke by 6mm from 81.4 to 87.4mm and bore (limited by the factory bore spacing) from 95.5 to 96.0mm, increasing total displacement from 3498cc to 3796cc, or 3.8 liters.
Increasing the stroke by 6mm has a noticeable effect on piston speeds-a significant concern for an NA engine designed for high-rev power. The original dimensions are over-square (larger bore than stroke) with a bore/stroke ratio of 1.17, while the 3.8-liter dimensions would bring the engine closer to square (equal bore and stroke) with a ratio of 1.10. Although the engine is still over-square, which is typically good for higher revs, the mean piston speed at 8500rpm has increased from 23.1m/s to 24.8m/s; 25 m/s is roughly F1 engine territory.
However, mean piston speed (or the average speed of a piston through one revolution) only helps classify the type of engine. Higher speeds usually mean higher performance. When changing an engine's internal geometry, such as adding stroke and changing rod lengths, what matters more is the piston velocity profile and the amount of sideways thrust added to the piston.
When an engine is stroked, the rod journals get moved further outboard from the crank centerline. The original VQ35DE has a stroke of 81.4mm, meaning the rod journal rotates 40.7mm (or half the stroke) from the crank center axis. This is because an engine's stroke is the total vertical distance the piston travels as the crank rotates 180 degrees between one and six o'clock, which translates exactly to the distance the rod journal has to travel. By adding 6mm of stroke, Cosworth had to increase the journal offset radius by 3mm. The added radius has a side effect, though, since it will shove the piston harder against the cylinder wall as the crank sweeps from one o'clock to five o'clock on the combustion stroke. This can increase wear, drag and may introduce ring flutter and piston wobble at high speeds.
Increasing the journal offset radius by 3mm also brings up another issue: now the piston will pop 3mm above the top of the block into the head at TDC. Since the VQ's deck height can't be changed, there are only two options to compensate for this. Either shorten the rod by 3mm, or modify the piston design and push the piston wrist pin position up by 3mm so the piston sits lower than stock. Here's where rod ratios come into play. Rod ratio is basically rod length divided by stroke length. Long rods or short strokes will have a large rod ratio (roughly 2:1), like high-revving engines such as the B16B or SR16DE. Short rods or long stroke will have a low rod ratio (approaching 1.5:1) like Nissan's 2.5-liter QR25. These can't spin fast, but have massive low-end torque. While rod ratio has implications on torque and engine speed, it's more an indication of how much side load the pistons are subjected to.
In our case, if rod length was decreased by 3mm to match the stock deck height, we would end up with a lower rod ratio of 1.61:1. Imagine the right triangle formed by rod, crank and piston position. A larger journal radius (which makes up the horizontal leg of the triangle) combined with a shorter rod (the diagonal leg) will increase the angle the rod forms to the cylinder wall. The larger the angle, the harder the piston shoves sideways into the cylinder wall. The increased rod angle would also require shorter piston skirts that might compound the issue of piston wobble.
In the VQ35DE, using a shorter rod to compensate for the stroke increase would be a double whammy. Instead, Cosworth chose to reduce the wrist pin depth and retain the stock rod length.
This gives a slightly higher rod ratio of 1.65, compared to the stock VQ35DE's 1.77, resulting in a peak piston speed increase of 9m/s (or, for ease of comprehension, 22mph). The only compromise here is that the ring packs also have to be moved up, reducing the space available for each ring land. Loss of ring land is bad, since it doesn't hold up as well under knock. But as we're running 100-octane gas under NA power, the trade-off is minimal. Small ring lands are also cleaner on emissions and could provide a fractional increase in power.
Compression ratio is only limited by the fuel used. And since our engine will deploy off-the-shelf, CNC-ported Cosworth cylinder heads-designed for both turbocharged or NA applications-our 12.1:1 compression ratio will come strictly from the high-compression Cosworth slugs thrown in. This ratio is based on previous experience, since combustion efficiency, chamber design, tuning and octane all interact to restrict how much compression an engine can run.
That takes care of our bottom end.
Next time, as we continue work on our Castrol Syntec Top Shop Challenge engine, we'll address all the less analytical and more black-art stuff requiring more real-world testing, like heads, ports, cams and tuning.