I’m in the planning stages for my Z4 3.0i build. After debating an engine swap or going turbo, I’ve decided a ~15 psi supercharged M54 would best meet my build’s goals of reliability, 400+ whp, and remaining daily drivable. The next decision for me to make is whether to stick with my Vortech, or move to an ESS twin screw.
Bottom line, I’m considering a high compression build to max out my current Vortech supercharger at around 15-16 psi and 45k-50k impeller rpm, adding an intercooler, and spraying a water/meth mix for detonation control. Alternatively, I could move to a TS3 setup, but with custom low compression pistons and rods. Tuning has already been lined up and is not an issue.
My thought process is outlined here, and constructive criticism is welcome. I wanted to do my own research privately and keep quiet about it until the build was complete, but I’ve found some pretty interesting things that I haven’t seen anywhere else… at least not on a BMW-centric forum.
It’s common knowledge that a twin screw makes boost sooner. Centrifugal superchargers seem to make more peak power at similar boost levels. Both twin screw and centrifugal supercharger companies claim that their supercharger is the most efficient out there. But who is right? What does the “adiabatic efficiency” really mean? Now that I can change compression ratio,
should I change it? And to what? How will gas mileage be affected?
I couldn’t find much real data out there other than forum myth, so I had to be creative.
http://www.mustangandfords.com/proje...ced-induction/
This is the only instance I’ve found of the same vehicle being equipped with various types of forced induction, run to 14-15 psi, and measured on the same dyno. Yes, it’s not an M54B30 and the compressors are different models than the ones I’m considering. But, I think generalizations can be made for the BEHAVIOR of each type of compressor in a high boost situation.
On the Mustang, boost builds with both compressors as rpm climbs. It starts higher with the twin screw, around 11.9 psi @ 3k rpm, and then climbs slowly to 14.5 psi. The centrifugal supercharger starts lower, around 2.2 psi @ 3k rpm, and climbs exponentially to 14.5 psi at 6600. For my build, I will likely set the rpm limiter to 6800 rpm. I extrapolated the twin screw running 14.5 psi to 6800, and the centrifugal increasing to 15.5 psi because I plan to run 15-16 psi. Since boost is a measure of restriction, it’s possible that the Mustang’s engine can’t flow enough air to support the supercharger, so it starts to choke up top, and boost pressure builds as flow requirements increase. A flow restriction would affect both superchargers, and without additional data showing supercharger behavior in other high boost applications, I am using this test as a generalization of supercharger behavior.
Predicting exact boost pressures wasn’t imperative to extract the I data I wanted, and everything would change anyway once I install aftermarket camshafts and have my cylinder head modified.
What the boost curves allowed me to do was plot performance on each supercharger’s compressor map. I used formulas to predict volumetric flow in cubic feet per minute (cfm).
The twin screw is plotted as a near-horizontal line with a slightly increasing slope, since the pressure ratio [(boost pressure + 14.7) / 14.7] is almost constant. The Vortech’s performance is plotted as a line with steep slope, which skirts the 70% efficiency line to the left of the supercharger’s “sweet spot”. For each data point, I found the compressor efficiency at 100 rpm increments. Yeah, it was time consuming to plot all 39 data points on each compressor map and transfer them to an Excel file :lol.
So, which is more efficient, the twin screw or centrifugal supercharger? Both. It depends on where you are on the compressor map. Below 4.2k rpm, the Lysholm is more efficient. Above 4.3k rpm, the Vortech is more efficient.
What does this mean?
Using some thermodynamics, you can predict the discharge temperature and horsepower lost due to compression.
For the discharge temperature, I used an inlet temperature of 70 deg F. As you can see, the Lysholm’s discharge temperature is a lot hotter, 224 deg F @ 3k rpm versus 111 deg F for the Vortech. Even though the twin screw has a better efficiency at 3k rpm, it produces more than 5 times as much boost, and the penalty is an additional 113 deg F and 5.1 hp loss over the Vortech.
These discharge temperatures are only accurate for the air leaving the supercharger. The air is then cooled with an intercooler/aftercooler and/or WMI. The ESS kit utilizes a liquid aftercooler. Water has four times the specific heat of air, meaning water can absorb four times the heat of air before it rises one degree. Coolant has a lower specific heat than water, so cooling capacity will vary depending on water/coolant mix. Besides the fact that air-to-air systems aren’t normally feasible for positive displacement superchargers, I think the air-to-water system is the right choice for the ESS kit, since its discharge requires a huge drop in temperature to avoid detonation. It also explains the reduction in compression ratio for the TS3 kit. However, I am concerned about heat soak during extended driving with the TS3 kit.
The Vortech may get away with an air-to-air intercooler, and water/meth for high discharge temps in the upper rpm band. Compression ratio can be higher due to the lower discharge temps and water/meth mix for detonation control.
Now to the big question - which one will be faster?
Obviously there are assumptions that must be made. I used a rule of thumb of 4% change in horsepower per point of compression. I calculated wheel horsepower in 100 rpm increments by taking a stock M54B30 dyno; multiplied it by the pressure ratio of each compressor; multiplied the result by % gained or lost due to CR; then subtracted the power it takes to spin the supercharger. I did not take into account power adders such as headers, cams, head work, and so on… nor did I take into account power lost due to heat of compression, since it is too dynamic.
With a 10.7 CR (102%) for the Vortech, and 9.0 CR (95%) for the Lysholm:
Below 5k rpm, the Lysholm makes an average of 30% more power than the Vortech, good for 50-66 more horsepower with my rough estimates. The Vortech finally moves ahead around 6.1k rpm, peaking about 11% higher at redline.
In a race with a Vortech versus the Lysholm, first gear would be a wash due to traction. Shifts at redline would land the cars at 4k rpm in 2nd; 4.5k in 3rd; 5k in 4th. Even in an ideal race, where both drivers shifted exactly at redline with no traction issues, the Lysholm car would be significantly faster until the top half of 4th gear.
Now if we raise the CR of the Vortech Z4 to 11.5:
The crossover point is lower (5800 rpm) and peak horsepower difference is 16% in favor of the Vortech. In this situation, I think the twin screw would still be faster through 2nd and 3rd gear. Torque wins races, and it’s tough to come back from a deficit.
It’s safe to conclude that with all other things being equal, a TS3 at 14-15 psi would be faster than a high compression Vortech running 15-16 psi and WMI until triple digit speeds. However, it would also consume 6 hp at highway speeds (and have a lower CR), while the raised compression of the Vortech Z4 would offset the power consumed by the supercharger (and the car would possibly get better gas mileage than stock). Additionally, the Vortech Z4 would require less cooling capacity, especially during normal driving.
If I had an ESS twin screw, I wouldn’t switch to a Vortech and build the motor to get the added top end and gas mileage. For me, it’s a matter of cost versus benefit. I already own a Vortech; I’ve got a nice trunk-mounted water/meth setup; I have a 3.46 Quaife LSD that might be overwhelmed by the extra low end of the twin screw; and I daily drive the car and would appreciate the added gas mileage. However… custom crank/accessory pulleys, injectors, intercooler, brackets, piping, etc. aren’t cheap, and there is a salvage value in selling my VF kit.
Both setups have pros and cons, some of these results surprised me, so I wanted to share
------------------------------------------
Update 2/13/2016
Before I start swapping pistons and rods, I wanted to see if there would be any benefit to using a smaller pulley on a centrifugal setup, and adding a wastegate between the supercharger discharge and throttle body to control boost levels. The engineer in me doesn’t like the idea of it because it’s an inefficient way of doing things. Regardless, I considered doing it to get max boost from ~6k rpm to redline, with the tradeoff of higher discharge temperatures and more power consumed by the supercharger.
In other words, I’d get more low end and midrange, at the expense of peak power and efficiency.
My tuner says he can tune around it. I think part throttle would be hard to tune around, because you have two competing discharges. The intake wastegate and bypass valve are both dumping boost (unless the bypass opening dropped boost pressure enough that the wastegate closed). The ECU is using the info given to it by the MAF sensor for fuel calculations, so the car would need to be tuned to estimate how much air is being dumped through the wastegate, and how much is being dumped through the bypass. At full throttle or off throttle, things are better defined.
For the estimations, I used the same boost curve that I used previously for the 14.5 psig twin screw and 15.5 psig centrifugal (henceforth, I’ll drop the “g” from psig). I multiplied each 100 rpm increment by a fraction related to the new peak boost (so in this case, I multiplied each data point by 8.5/14.5 for the twin screw and [8.5, 12, 15, 18]/15.5 for the centrifugal). I did not go back and recalculate the compressor efficiency at every single data point. I did glance at the compressor maps, and the twin screw operates just above the “sweet spot” at 14.5 psi, and just below it at 8.5 psi; the centrifugal starts out in an area not covered on the compressor map (so data would need to be extrapolated), then moves towards the “sweet spot” of 73% efficiency. For the purposes of what I’m showing, a 5% change in compressor efficiency is irrelevant, since it only equates to a few degrees of discharge temperature and tenths of a horsepower.
The 15 psi and 18 psi pulleys would probably require a whole new pulley set to avoid belt slip, since the supercharger pulley would be too small with the standard crankshaft pulley, and the power required to turn it is higher. A Vortech SCi trim compressor would need to be spun to about 53k rpm (2.36” pulley using the stock 5.11” crank pulley… and that won’t happen without serious belt slip :lol).
First is boost pressure. Peak boost is reach at:
6800 rpm, 8.5 psi pulley
5900 rpm, 12 psi pulley
5300 rpm, 15 psi pulley
4900 rpm, 18 psi pulley
Even with maxing out the centrifugal supercharger, the twin screw has a huge advantage at low rpm.
Next we have the volume flow rate (cfm) of air dumped by the wastegate, and air passing through the throttle body. Ideally, only the air that is passing through the throttle body would have passed through the intercooler. The air that is not going to be used by the engine would be dumped before it passed through an intercooler. All things being equal, a lower volumetric flow rate means cooler air will exit the intercooler and into the engine, because there is less hot air to cool.
If the boost pressure graph shows the advantages of smaller pulleys, the discharge temperature graph is one of the areas where the downsides appear. I used an ambient temperature of 70 deg F, like with my previous high boost analysis (by the way, “standard day” is actually 59 deg F).
Obviously, some sort of aftercooling is necessary. Max discharge temps for each setup:
174 deg F, 8.5 psi
209 deg F, 12 psi
236 deg F, 15 psi
261 deg F, 18 psi
196 deg F, twin screw
When I lived in the California desert, ambient temperatures regularly got up to 110 deg F.
222 deg F, 8.5 psi
259 deg F, 12 psi
289 deg F, 15 psi
316 deg F, 18 psi
246 deg F, twin screw
I think we can agree that a well-developed means of cooling is necessary.
A smaller pulley is a compromise. It provides the midrange benefits of a high boost setup without the stress of high boost pressures, which is great for a stock engine. Because the volumetric flow rate of air entering the intercooler is lower, less cooling capacity is required, even if discharge temperatures are the same.
Then again, you deal with many of the downsides of a high boost setup, without enjoying the significant power advantages that high boost provides. The power required to spin the supercharger is the same whether you use the boost or dump it, so there’s a lot of waste:
Using the formulas I used previously, we can estimate wheel horsepower. Again, this doesn’t take into account power adders such as headers, FI-spec camshafts, or head work. It also doesn’t include losses due to intercoolers, heat of the intake air, or conservative tuning.
IMHO, a 12 psi pulley, with an 8.5 psi wastegate dump, appears to be the best compromise for a centrifugal setup. Max boost comes in before 6k rpm. Discharge temperature and horsepower consumed by the supercharger are mildly greater than a twin screw at high rpm, and it will run cooler and be more economical everywhere else. However, for maximum average horsepower at identical peak boost levels, you can’t beat a twin screw.
Overall average horsepower, 3k-6.8k rpm:
226 whp, 8.5 psi pulley
240 whp, 12 psi pulley
247 whp, 15 psi pulley
250 whp, 18 psi pulley
264 whp, 8.5 psi twin screw
Average horsepower, 4k-5.5k rpm:
216 whp, 8.5 psi pulley
233 whp, 12 psi pulley
247 whp, 15 psi pulley
255 whp, 18 psi pulley
263 whp, 8.5 psi twin screw
Average horsepower, >5.5k rpm:
281 whp, 8.5 psi pulley
297 whp, 12 psi pulley
295 whp, 15 psi pulley
291 whp, 18 psi pulley
298 whp, 8.5 psi twin screw
If you’ve got a centrifugal supercharger, it may be worth pursuing this mod, along with aggressive gearing to minimize your time spent outside of boost.