How to fight exhaust reversion when the primaries are way too short

[StockerChevy]

Some of the FAST racers under cam a big compression motor with wider LSA and retard the camshaft a considerable amount. You would be surprised at the outputs and rpm band the motor will work at. The intake valve has to flow well at low lifts. There are other tricks as well but a big cubed Hemi racer runs 840 horsepower through dual exhaust 2.5 inch pipes through decent exhaust manifolds.

Jim

[ptuomov]

Some of the FAST racers under cam a big compression motor with wider LSA and retard the camshaft a considerable amount. You would be surprised at the outputs and rpm band the motor will work at. The intake valve has to flow well at low lifts. There are other tricks as well but a big cubed Hemi racer runs 840 horsepower through dual exhaust 2.5 inch pipes through decent exhaust manifolds.

Jim

Yes, please! More information on those top FAST motors. That's precisely the type of info I'm looking for, as they all have the same exhaust blowdown interference problem that I have.

[modok]

being four valves per cylinder makes it a lot easier. I bet you can mess with the cam timing a lot if needed.
Get a pressure trace on it and see how bad the problem actually is.

[Calypso]

A resonator tube is going to reflect a positive pressure wave back to #1 and #3. How do you reflect a negative wave to #1 in that manifold? (I've got a turbine corking the manifold outlet.)

Maybe I’m way off, but wouldn’t 1/4 wavelength resonator connected to #3 be phase shifted to cancel out the wave action at the designed frequency. Of course, if the pressure in the manifold is higher than in #1, flow will still follow, but less in tune with that rpm. Thinking a dead end channel in the manifold.

It's my impression that in a logish manifold like this, each exhaust blowdown event is so violent that it effectively resets the whole system. Blowdown events happen as frequently as 400 times per second at 6000 rpm, since 90 degrees of crankshaft movement takes 1/(4*6000/60) = 0.0025 seconds. The speed of sound is of the order of 2000 ft/s in that temperature, so you have time to move about 5 feet (more in the direction of flow, much less against the flow) before the new "reset" pulse comes in 0.0025 seconds. I don't see how I could get a quarter-wave resonator to do anything beneficial in that sort of setting, can you sketch out your idea in more detail? Or have I misunderstood how it works?

Do I understand correctly that your thinking is that each arriving pulse is so strong that it would kill previous wave action therefore rendering resonator useless, because it’s effectively a delay line phase shifting previous pulse at the resonant frequency and it wouldn’t work in single pulse environment, because of the delay?

[modok]

I agree.
Wave action is not significant with a short manifold and turbo.

[cjperformance]

Nitro2

[mag2555]

I would apply a valve job to the Exh valves and seats in the problem cylinders that inhance low lift flow to the max.

[ptuomov]

I agree. Wave action is not significant with a short manifold and turbo.

This issue is entirely about wave action. It's the blowdown wave that prevents the evaucation. The flow never goes in the wrong direction.

Nitro2

I can't get the pressure sensors to fit easily on this engine that is in the car already (for multiple reasons, including the need for a water cooled sink). I'll have them on the next engine. But in this case, I think I already fully know what the problem is. I just don't know how to solve it, given the constraints.

One potential mitigant is retarding the cams and that may help normally aspirated car as it makes the exhaust valve open later and exhaust gas a little cooler. For a turbo car, I don't want to open the exhaust valve too late. I need to open it pretty early to spool the turbine at low rpms, so this trick is not applicable.

Another NA trick that might work is increasing compression. Since the exhaust manifold isn't going to evacuate the combustion chamber, use the piston to do that. Small chamber volume will make it possible to just use the piston to push the exhaust gas out. Unfortunately, small chamber volume and high compression ratio is also out of question in a turbo motor.

I would apply a valve job to the Exh valves and seats in the problem cylinders that inhance low lift flow to the max.

To which cylinders and why?

[gottago]

re...

Unread post by StockerChevy » Thu Nov 16, 2017 1:42 pm

Some of the FAST racers under cam a big compression motor with wider LSA and retard the camshaft a considerable amount. You would be surprised at the outputs and rpm band the motor will work at. The intake valve has to flow well at low lifts. There are other tricks as well but a big cubed Hemi racer runs 840 horsepower through dual exhaust 2.5 inch pipes through decent exhaust manifolds.

Jim

I went a similar route as this to some degree recently in a na application. Ended up there by way of the cam drill. Have to agree it was a bit surprising.. At times this sort of n/a combo almost reminds you of being in a turbo car. Great power band. So I got to wondering about how to incorporate a turbo to expand on this. I suppose its just cam events in relation to compression and rpm that changed but reversion issues all but disappeared along with and it just seemed to have potential for more.. I was using a wider lsa cam with same in and ex duration. 112 icl. Maybe StokerChevy has some ideas on how to apply this sort of thing to your situation. He did offer this up in regard your post.

[ptuomov]

re...

Unread post by StockerChevy » Thu Nov 16, 2017 1:42 pm

Some of the FAST racers under cam a big compression motor with wider LSA and retard the camshaft a considerable amount. You would be surprised at the outputs and rpm band the motor will work at. The intake valve has to flow well at low lifts. There are other tricks as well but a big cubed Hemi racer runs 840 horsepower through dual exhaust 2.5 inch pipes through decent exhaust manifolds.

Jim

I went a similar route as this to some degree recently in a na application. Ended up there by way of the cam drill. Have to agree it was a bit surprising.. At times this sort of n/a combo almost reminds you of being in a turbo car. Great power band. So I got to wondering about how to incorporate a turbo to expand on this. I suppose its just cam events in relation to compression and rpm that changed but reversion issues all but disappeared along with and it just seemed to have potential for more.. I was using a wider lsa cam with same in and ex duration. 112 icl. Maybe StokerChevy has some ideas on how to apply this sort of thing to your situation. He did offer this up in regard your post.

Thank you, gottago. Also thanks to StockerChevy, form whom I've got a lot of ideas.

I think the main differences between the turbo and normally aspirated engines here are the following.

First, to spool at low rpms, early EVO is preferred in a turbo car. (With the latest set of cams that have earlier EVOs, the turbos spin 5000rpm at idle.) In a NA car, you don't pay a similar low-rpm penalty for late EVO. I have feeling (not knowledge) that earlier EVO and slightly larger turbine is superior at high rpms to later EVO and slightly smaller turbine, if you want to keep the boost threshold rpm constant.

Second, high compression helps with a NA car, because that allows one to use the piston to forcibly pump more of the exhaust out against whatever back pressure. But a turbo car needs a low compression to make a lot of power with pump gas. So somehow one needs the exhaust wave to evacuate the chamber.

Third, run the engine at such high rpms that the 90-degree exhaust interference wave (that arrives before the victim cylinders intake valve opens at 5000 rpm and that arrives during the victim cylinder's overlap at 6000rpm) arrives after the exhaust valve has already closed. That is, let the engine run poorly at 6000 rpm (lower the boost and retard ignition there) and focus on making power at 8000rpm. This would mean changing basically every rotating and reciprocating component in the base engine. Without variable valve timing, it might also compromise the low rpm drivability. This is the solution of last resort, given the development effort and uncertain success.

[user-23911]

First, to spool at low rpms, early EVO is preferred in a turbo car. (With the latest set of cams that have earlier EVOs, the turbos spin 5000rpm at idle.) In a NA car, you don't pay a similar low-rpm penalty for late EVO. I have feeling (not knowledge) that earlier EVO and slightly larger turbine is superior at high rpms to later EVO and slightly smaller turbine, if you want to keep the boost threshold rpm constant.

A shorter cam works best to give low RPM torque.
A longer cam works best to give mid and high RPM torque.
A smaller turbine works best with low RPM torque.
A bigger turbine works best with mid and high RPM torque.


From that a small turbine works best with a short cam and a gig turbine works best with a long cam.

If a turbine is spinning 5000 RPM at idle then it's not tuned very well. Idle speed is the key point.
Any engine running well that idles at 750 or so RPM, the turbine won't spin or if it does, it'll spin so slow you can see it.



Just take a look at some factory turbo cam specs and they're usually the same as an N/A spec give or take 2 or 3 degrees.

[ptuomov]

If a turbine is spinning 5000 RPM at idle then it's not tuned very well. Idle speed is the key point.
Any engine running well that idles at 750 or so RPM, the turbine won't spin or if it does, it'll spin so slow you can see it.

The engine idles at 750 rpm and drives beautifully. The rpm reading is from the turbo speed sensor, admittedly not a lot resolution there with that 5000rpm. But we can see the turbines spinning, not furiously but clearly spinning, at idle when the exhaust is off. They started spinning when we switched to earlier EVO cams (and slightly larger turbines).

[Erland Cox]

I run as short manifolds as possible with a turbo.
That way I can really use thepulse loading instead of the flow.
It is like knocking someone instead of pushing.
Much better spool up to.
I do not use very much more lobe separation than in an N/A engine.
I prefere pulse split in the exhaust, it makes everything easier.
In a 4 cylinder that is 1-4 and 2-3.
But I have used extremely short exhausts with 1-2 and 2-3 into the turbo.

Erland

[ptuomov]

I run as short manifolds as possible with a turbo.
That way I can really use thepulse loading instead of the flow.
It is like knocking someone instead of pushing.
Much better spool up to.
I do not use very much more lobe separation than in an N/A engine.
I prefere pulse split in the exhaust, it makes everything easier.
In a 4 cylinder that is 1-4 and 2-3.
But I have used extremely short exhausts with 1-2 and 2-3 into the turbo.

Erland

That’s not a cross plane V8, is it?

[modok]

I agree. Wave action is not significant with a short manifold and turbo.

This issue is entirely about wave action. It's the blowdown wave that prevents the evaucation. The flow never goes in the wrong direction.

Waves are superimposed on the main flow. Waves are things happening in addition to the main flow. It's just flow. There is no monkey motion hiding within it. If goes in one end, it comes out the other. There are waves, certianly, but they are too short to do anything with them. The whining of a basic supercharger is waves, but you don't see anybody trying to harness them