HiWire

I know lots of people have written about the many weaknesses of the sequential turbo setup (as Mazda implemented it in the FD), but what are the chances they will use it again? Is there another, more modern way of designing sequential rotary turbos now, especially with more advanced engine management, variable-geometry turbos, and direct injection?

I'm reading this fantastic RX-7 vs. Supra thread for reference: https://www.rx7club.com/3rd-generati...turbos-960727/

scotty305

iTrader: (2)

Assuming two small turbochargers cost approximately twice as much as one large one, I would be surprised if this happened.

This isn't exactly related but at one of the SevenStock events I attended, I had the good fortune of chatting with Mr. Takaharu Kobayakawa who was involved in the FD project. I explained that I own and enjoy one of the cars he was responsible, and asked him if he would have changed anything about the FD knowing what he knows today. I don't remember the exact words, but he said he would have used something like the Renesis engine if it had been available at the time. He was of the opinion that the FD would have been more successful if the cost, weight, and complexity had been decreased even if that meant sacrificing a little horsepower. Personally I think I'd like a little more power than a non-turbo 2-rotor can offer but it would be interesting to see how an FD might feel with ~200whp, zero turbo lag, and about 100lbs less weight on the front of the car.

HiWire

Interesting. You could say that Kobayakawa got what he wanted with the RX-8, then, but with more chassis weight.

A friend sent me a link regarding Flying Miata's V8 car... I think I'd be more interested in seeing a Renesis transplant in a sub-2,500 lb. 3rd-gen MX-5.

DriftDreamzSS

iTrader: (4)

I would think the advancements in turbocharger technology, ecm's and the variable geometry turbos would pretty much negate the need for a sequential setup. I think with today's technology you can have a much broader and usable powerband with a single turbo setup than you could have in the early 90's.

HiWire

Yes, it would probably make for a cleaner engine bay and allow more room for cooling, etc. as well.

It's interesting that Mazda didn't "market" the turbo technology with a badge on the FD, like other manufacturers. I've always found that kind of labelling gauche and dated.

t-von

iTrader: (8)

The sequential system was designed to help add torque in the very low rpm ranges of a small displacement engine. The new 16x is suppose to have a decent torque increase in NA form making using such a system pointless. At this point, Mazda is gonna everything in their power to make sure the next rotary is simple and reliable.

WJM ROTARIES

They should use something like a gt35 on it and get the best of both worlds ,no lag and heaps of torque and reliable

arghx

iTrader: (3)

So this is an interesting topic because sequential turbos are now common on passenger car diesels, and gasoline engines are slowly adopting technology from diesels such as direct injection and EGR coolers. The sequential turbos used on diesels can also be referred to as "compound turbos." Here's an image that shows the modern way sequential turbos are controlled hardware-wise:



So you've got a small turbo and a big turbo. At low speed the exhaust is directed at the small turbo first. The fresh air gets sucked through the big turbo, then compressed where it is forced through the small turbo.

At mid rpms the exhaust flow is split between the small and large turbo. At high rpms the exhaust flow goes right to the big turbo, and the compressor side of the big turbo is allowed to flow directly into the engine instead of going into the small turbo first.

Now let's take a look at something else. I posted this in another thread. It's a chart I made which compares the FD torque curve (as reported by Mazda, at the flywheel) to the most state of the art high revving turbo direct injected engine on the market today, an AMG 2.0 liter engine.



You can see that the AMG engine, which is a single turbo, still has higher lower speed torque than the FD as measured on an engine dyno at the flywheel. That's what 20+ years of development will do. So Mazda has a long way to go to compete with the very high end turbo engines.

The thing is, a powerband can only be so wide with one exhaust-driven turbo. That's where two turbos come into play, or a conventional turbo and some other boosting device (mechanical supercharger or something electric). A GT35R style powerband is way too laggy for regular customers who are buying a brand new car with a warranty. It also takes a very open intake and exhaust for a GT35 to make the power it does, which won't meet noise and emissions requirements for OEMs.

wstrohm

arghx,

Thanks much for those diagrams... looks like a very neat setup.

HiWire

Yes, thank you!

I wonder what is in the AMG 2.0L turbo engine that makes so much power at the low end (and overall) – I assume that it is running at high boost pressure. My friends and I saw the CLA45 at a show this winter and thought the power figures were astounding for the displacement (though the car was also expensive). It is also amazing that it is emissions-legal.

With technology like this, it's no wonder that Aston Martin is planning to use Mercedes-AMG engines.

arghx

iTrader: (3)

The AMG torque is terrible at the low end by modern DI standards. That's due mostly to the size of the hotside on the turbo. Current 2 liter engines are doing closer to 350nm. Take a look at the 2.0 LTG engine from GM for the Cadillac ATS:



The main thing that drives low speed torque on modern turbo DI engines, besides displacement of the engine, is the sizing of the turbo and the turbine power generated by the exhaust.

The sizing of the turbo on the low end is driven by two constraints: compressor surge and turbine A/R. The compressor surge is determined by the shape of the compressor map. If the compressor is geared towards high flow at peak power, or a lot of performance at high altitude, the engine will have compressor surge at low speed and high load. This is difficult to experience in a car, but you can run into it if you are lugging in top gear at low speed up a grade.

There are tricks to make the compressor map wider but they all hurt peak efficiency (change compressor trim, change diffuser design, etc). If you hurt peak efficiency your compressor outlet temperatures will rise and now you've put more load on your intercooler and more risk of knock.

The turbine power itself is determined primarily by the mass airflow through the engine and the temperature of the exhaust. The mass airflow through the engine will be determined by the gas exchange events (valves on piston engine, ports on rotary). Typically you can increase mass flow by incorporating low speed scavenging from overlap. However that scavenging can cause emissions problems by creating NOx tailpipe emissions or heating up the cat by causing exothermic reactions. The emissions aspect is not a big deal in the aftermarket but it will cause headaches for the OEM trying to develop a new product. The temperature of the exhaust can be increased through spark retard to speed up the turbo, but that can make the combustion unstable resulting torque surges.

The turbine power required to make boost is determined by the turbine A/R. A lower A/R means less turbine power needed, but it also means more restriction at peak power. More restriction at peak power means more backpressure for the engine to work against, and more residual gases that can create knock.

Flybye

The real problem is our system is overly complicated. Look at the MKIV's sequential diagram:



Yeah that's it! I dont know why Mazda had to make it so crazy complex. Hell's no I wouldnt put in a NA renesis in the FD. Its nice to cruise at 3k rpm around 70ish, give it a little gas to pass someone, I get instant boost, and not have to worry about downshifting if I had something in my hand.

The rats nest could have been a lost simpler, but it wasn't. But oh well.

arghx

iTrader: (3)

I speculated in my article that the FD system is complicated because it used swing valves instead of butterfly valves. The reason for that is space constraints.

A lot of the FD rats nest also has to do with the poor combustion requiring the elaborate smog pump system control, and also various actuators needed because it didn't have an electronic throttle.