PDA

View Full Version : Two stage twin-scroll turbocharger



DannytheRadomski
05-27-2014, 08:53 PM
I have been reading up on twin-scroll turbochargers. I was wondering if you could do it backwards with the multiple scrolls in the compressor housing instead of the turbine housing. My thought process behind this is to have a two stage turbo or variable geometry turbo effect (with boost at lower and then higher rpms) by having a small scroll providing boost early and then a bigger scroll in the housing providing a more powerful boost. I don't know if this would work or not with the current design of turbochargers, but I feel like it could be a very simple and effective advance on turbocharger technology.

Any responses are helpful and appreciated, whether they support my idea or not.

Jay Lawrence
05-27-2014, 10:46 PM
Danny,

Are you thinking of 2 different A/R ratios on one compressor housing? If so, I imagine this would cause the lower backpressure path on the exducer to be favoured, so you'd end up with most of the air flow going through that. If both paths were subject to equal back pressure, then I imagine the smaller A/R (with it's higher outlet pressure) would try to overcome the larger A/R (with it's increased boost response) and you'd end up with a net, errr, something not great. Also keep in mind that A/R has much less of an effect on a compressor than on a turbine, and that the twin scroll turbine is typically fed per cylinder pairing (I don't think feeding a cylinder pairing with a twin scroll compressor would be a good idea (different boost levels in different cylinder pairs)).

DustinGC8
05-29-2014, 05:46 PM
Jay is correct. The twin scroll turbine housing reduces spool times by removing interference between exhaust pulses, not by one scroll acting on a smaller section of the turbine wheel. This has the added benefit of reducing back pressure in a properly tuned exhaust manifold by creating a vacuum effect between paired cylinders. An example would be the 1-3-4-2 firing order of most inline 4 engines. Cylinders 1 and 4 are paired to one side of the turbine and 2 and 3 are paired to the other. In this setup, no two exhaust pulses will ever collide with each other and each pulse serves to draw out the next one behind it.

The situation you are describing in your post is best accomplished using a traditional variable geometry turbo or a complex sequential twin setup that would use two different turbo sizes.

-Dustin

DannytheRadomski
05-30-2014, 07:15 AM
I know how a twin scroll turbo works, and I know the benefits of it. I am not saying use the same thing and expect different results; I am proposing a new idea. I would like to hear about whether or not my idea is possible, not that a traditional twin scroll turbo is different from my idea.

Jan_Dressler
06-01-2014, 06:16 AM
I would like to hear about whether or not my idea is possible
Jay already answered to this, and I am thinking the same: No, IMHO it won't work, because of the reasons Jay mentioned.
You would basically need two different sized compressors on the same shaft, and some kind of a valve system to "deactivate" the bigger one. And, to use it to full extent, also two different turbines, or a VGT. Which would mean a physically large and heavy turbo with a high inertia (slow spool-up), plus the valves. At this point it would be better to use the sequential twin turbo setup that Dustin mentions. Or build your own electric assisted turbo ;)

VFR750R
06-06-2014, 08:45 PM
I am under the impression reading several books on turbos, google etc that turbochargers generally do not respond (are not sensitive) to A/R ratio on the compressor side. Most of the sensitivity is related to the inducer diameter and the wheel itself. All of the 'compression' occurs in the wheel, the compressor cover just seals against backflow as best as possible and gives an area to collect the outlet gas with reasonable efficiency and direct it to a single outlet.

Certainly, like anything else regarding turbochargers, compressor covers are probably compromised to cover a range of operating conditions even though turbos would prefer to be steady state devices, I do not believe it's an area that would result in gains commensurate with effort.