What happens when a divided turbo setup isn't "fully" divided?

For anyone curious what can happen to boost response using a single wastegate on a divided manifold/housing, with and without a fully divided single wastegate provision. Only applies to setups that can benefit from a divided configuration to begin with (4cyl, 2 rotor, 6cyl) and to varying degrees (6cyl perhaps less so).

4cyl example, 2.0 4G63 with a divided T3 HX35 turbo setup, using a single 38mm wastegate.

The wastegate provision feeds from both cylinder pairs/scrolls, but is separated all the way up to the wastegate valve, with the divider matching the valve face shape to minimize any leakage/cross talk between cylinders. Not as “fully” divided as using two separate wastegates, but pretty good.

One pull is operating normally and the other has about a 10mm gap between the end of the divider and face of the wastegate valve. This gap opens a small path for cross talk and blowdown interference between cylinder pairs 1&4 and 2&3.

Both pulls are overlaid starting at WOT from a steady state 2500rpm, and marked at the nearest boost level. Same road, same direction, same temps etc. All done within a 45 minute window.

The normal configuration reached 27.5psi at 3644rpm after 5.78 seconds in 4th gear.
The gap configuration reached 26psi at 3952rpm after 7.56 seconds in 4th gear.

Four runs were done total, with the average difference in rpm and time being 285rpm and 1.75 seconds to achieve the same boost level. This was being conservative since the nearest available boost points in each log weren’t exactly the same (like the example above).

The difference between the two configurations was perceivable even when just moderately getting up to speed. The AFR changed slightly between the pulls, with the normal configuration being ~mid 12AFR during spool up vs ~high 11AFR with the cross talk. Dialing that in may slightly close the gap, but it’s really just a consequence of blowdown interference hurting volumetric efficiency (VE) during spool up, not the other way around.

More blowdown interference->more residual exhaust gas in the cylinder->less room for intake air/fuel to fill the cylinder (VE)->less overall airflow to spool/drive the turbo since the engine acts like it has less displacement. This effect is common and normal, and I'd argue it's the primary factor that allows a divided setup to improve spool/response from what we've seen on a 4cyl.



Ultimately, this shows the measurable change in boost response and boost threshold even a small opportunity for cross talk can have. It was still divided/separated throughout most of the wastegate provision, and the area for potential cross talk wasn’t even one square inch (roughly 0.6in^2).

If a single wastegate is going to be used on a divided setup, it’s important that the “divider” extends all the way to the wastegate valve face, which requires it to stick out PAST the wastegate flange.
I hate to say it, but designs that don’t will behave similar to what was seen here. However, some 6 cyl cases don't seem to be as sensitive, and we don't have direct experience with rotaries (they do share the same 180deg blowdown interference as a 4cyl though).

Does it still work if you don’t? Yes. But you’ll have a setup with boost response somewhere between open and divided, potentially losing much of the benefits of a divided setup when it comes to spool and response.

It's not necessarily bad, it just means there is room for improvement. Minimal impact otherwise (at full boost, peak power, etc.)