total induction volume with supercharger

[Mark911]

Needs some feedback. I've got a Acura NSX with an Eaton 1900 TVS supercharger. I've recently redesigned the induction system to add a much larger intercooler (air to water). This required much longer intake ducting and since the throttle body is in front of the blower has significantly increased the total induction volume that sees vacuum (although I still have the stock intake manifold at the "end of the line"). I'm afraid that all this volume will create sluggish off idle response as the energy to get all that air moving must be much higher now.

Here's where I need some feedback. I was considering adding a second throttle body right in front of the stock intake and connecting the two throttle plates together with a split throttle cable. The idea being that I could adjust the blower throttle plate to present near atmospheric pressure to the second throttle plate right before the stock plenum. This would allow all the parts of the induction system (intercooler and all associated tubes etc) to react much faster to minor (non boosted) changes in throttle position and help reduce any "lag" due the extra ductwork. I suspect I could simply open the blower throttle plates at idle until I no longer see any vacuum in the ducting prior to the newly added throttle at the intake manifold, then sync up the throttle cables to open both plates simultaneously.

Sounds too easy. What do you think?

Mark

http://www.mscperformance.com.

 

[MikeHalloran]

>>>. I'm afraid that all this volume will create sluggish off idle response as the energy to get all that air moving must be much higher now. <<<

You are proposing to add cost and complexity to solve a problem that you have only conjectured to exist.



Mike Halloran
Pembroke Pines, FL, USA

 

[Lou Scannon]

I believe the reason for locating the original throttle upstream of the blower is to minimize part throttle pumping losses. The tandem throttle arrangement you propose is used on blown drag race engines, if I'm not mistaken, and perhaps for related reasons.
The problem you have identified is sometimes called throttle "gulp", or "manifold filling effects", and has many aspects besides simply throttle response, when metering air and fuel flow are considered. Before I made any drastic changes to a factory setup like that you describe, I would want to understand how the air and fuel metering are realized, from a dynamic perspective, and what effects the proposed changes might have on dynamic metering. If you're lucky, your system is speed density, and port injected, which should be pretty immune to changes upstream of the intake manifold, assuming the changes do not appreciably affect reheat downstream of the measured or inferred charge temperature.
At any rate, drastically increasing the plenum volume between the throttle and intake ports will improve top end flow, if there is any improvement to be had from increasing plenum volume, but will have a commensurate negative effect on throttle response, due to the increased time lag between movement of the throttle and change of charge density at the entrance to the intake ports. The sensitivity of this time lag to plenum volume is the reason it is important to consider the dynamic air and fuel metering strategies when making such a change.
As MikeHalloran points out, the above mentioned potential issues may not be realized in your proposed modification. I also would add, engineering a tandem throttle setup as you propose will probably turn into a bit of a science project, before the desired outcome is achieved.

"Schiefgehen will, was schiefgehen kann" - das Murphygesetz

 

[Mark911]

My acura NSX is multiport EFI using a MAP sensor so yes, it's pretty insensitive to upstream events. I guess my original question was more theoretical in nature. Basically, would two throttle bodies (if designed and implemented correctly) solve the problem of sluggish, non boosted, transient throttle response on an engine with very large (relatively speaking) intake volume. I define intake volume as the total volume of the following in order from port to atmosphere; 1) intake manifold plenum, 2) runner to intercooler, 3) intercooler, 4) runner to blower, 5) runner to throttle body. Other than removing (actually relocating) the stock throttle body as is typical design practice for an externally mounted blower modification, the intake manifold is of stock volume. As stated by author Hemi, the relocation of the TB in front of the blower is necessary to avoid the power losses and other unwanted effects, which would occur if the inlet of the S/Cer were open to atmosphere looking at a downstream TB during low load operation. Mainly, the HP required to spin the blower which will produce excessively high pressures between the SC and a CLOSED (relatively) throttle plate. Obviously, the result would be a big waste of energy and unnecessarily high mechanical and thermal stresses at anything other than WOT.

Back to the theory. Since the dimensions of the stock intake plenum is unchanged, the mass flow and 2nd order acoustical tuning characteristics shouldn’t change too much. It’s the much-increased volume from the intake plenum to the relocated TB that is of concern. I realize that this volume produces a 3rd order acoustical effect as well. I agree that at WOT the stock intake combined with this larger volume might provide some top end performance gain. However, my main concern is off idle transient response, not WOT. As stated and as realized under actual driving conditions, there is a significant transient lag due to the increased kinetic energy necessary to first fill the huge vacuum space and then to get everything accelerated. Also as stated, this produces a fuel metering problem as well as the ECU sees a delta change in throttle position and load and responses accordingly with accel fuel. However, in fact very little air has actually moved.

I believe my proposed two TB solution would at least solve the first issue and go a long way to solving the second issue by providing a “atmospheric like” condition at the second (downstream) TB. In effect, the engine wouldn’t even know there was a SC upstream of the second TB, and that should provide as close to “stock” transient response as I can ask for at the expense of a very small amount of SC drive power.

At least that’s the theory as I see it. I’m looking for comments to either confirm my thought process or steer me in the right direction.

 

[LionelHutz]

Why can't the throttle body just be installed on the original intake manifold?

 

[BrianPetersen]

^ because the positive-displacement supercharger upstream of said throttle body would be trying to force air through the throttle even if said throttle were shut. Using a bypass valve around the supercharger is possible but the throttle response characteristics might be a bit odd (or need a lot of calibration effort to get it to work properly). The traditional Roots blower arrangement has the blower downstream of the throttles.

 

[timiacobucci]

Sorry for the delay, I'm not sure if you are still considering this but it is something I have been researching recently myself. Basically my understanding is that your thinking is exactly correct on about every point. This is actually a modification that has been well covered in the miata community.

http://forum.miata.net/vb/showthread.php?t=394583&page=2

It does solve the problems as you suspect very well and fine tuning the linkage between the throttles and pre blower stop point is very critical in how it functions. Depending on your boost levels you also may need a turbo style blow off valve in addition to the internal bypass if the 1900 has one, I know most other eatons do.


W

hy can't the throttle body just be installed on the original intake manifold?

You can't really run the blower unrestricted with a downstream throttle. It is not like a turbo, it is making boost all the time even at idle, from my reading even if you have the pre blower throttle open too far the blower will SCREAM all the time at idle and part throttle, aside from being unbearably loud it is inefficient and will wear the blower out prematurely as it is compressing and recirculating in a closed system at a fairly poor efficiency. It gets really hot really quickly and the lobes and housing can contact and damage the blower if it overheats too badly.

For this same reason, if you do have an internal bypass and you are fabricating the intercooler routing, it would be beneficial to route the outlet of the internal bypass after the intercooler and before the downstream throttle to deal with the heat it's generating while looping like that at low throttle opening.

Using a bypass valve around the supercharger is possible but the throttle response characteristics might be a bit odd (or need a lot of calibration effort to get it to work properly). The traditional Roots blower arrangement has the blower downstream of the throttles.

Most all modern supercharger systems have a bypass and you are correct that tuning it's activation is always an integral part of the how the system acts.

 

[JayMaechtlen]

maybe an actuator to close the throttle blade upstream of the blower based on manifold pressure?

What if you just let the blower build pressure proportional to the downstream throttle opening?
(regulate upstream throttle according to delta-P across downstream throttle?)

You might accept the pumping losses to have say 5 psi across the downstream throttle all the time, up to full throttle downstream where you'd just open the upstream throttle up fully.



Jay Maechtlen

http://www.laserpubs.com/techcomm

 

[timiacobucci]

I think you have the right idea and the principles at work figured out but you may be over complicating the control mechanism

You are talking about regulating the upstream throttle based on the pressure differential of the 2nd throttle. A direct 1:1 mechanical linkage already does this. The blower is a positive displacement device, it's output is directly linked to engine speed. When I said tuning the linkages I should have explained this better. I simply meant make sure it is 1:1. It is pretty common to make the pre blower throttle bigger than the 2nd so you need to size the linkage accordingly if the pulleys are different sizes for different throttles. If you are using 2 of the exact same throttle this shouldn't be an issue obviously.

From what I have learned 5 psi constant would probably destroy the blower, that's allot of air being recirculated at a very poor efficiency. It would get hot fast. The pre blower throttle stop is the constant pressure adjustment, even the slightest amount of positive pressure waiting on the other side of the stock location throttle makes the responsiveness of this system very immediate.

 

[JayMaechtlen]

If you regulate per pressure differential, would you need recirculation at all?
I really don't know if the 1:1 (or an fixed ratio) would do what you need - maybe it would.
anyway, hope the OP lets us know what he tries and what the results are.

cheers
Jay



Jay Maechtlen

http://www.laserpubs.com/techcomm

 

[timiacobucci]

Have you looked at the thread I linked to? There are actually many people already successfully doing this. These things are not really hypothetical at this point, it's been done as I described and it works very well.

Most modern blowers use a bypass for efficiency reasons at idle and cruise regardless of having no throttle downstream. It is needed though because it is always pumping even at idle unless it is a clutch driven supercharger, which some are. The compressed air needs to go somewhere if there's a closed throttle plate after it. A very sensitive blow off valve could be used to just vent it to atmosphere instead of recirculating perhaps if the blower you are using does not have any easy provision for one and your intention was to use it with a dual throttle system.

 

[timiacobucci]

This is a subject and experiment I am doing on my own. I wouldn't mind running some simulations but I also don't have the software for that at my disposal. Do you and have you learned something relevant about this subject?

Ultimately I think my time is better invested in just actually building it and data logging. Like you said nothing like real world data. The tvis and gsxr low end gains are such real data. What else explains this other than port velocity increases by reducing runner area? Intake manifolding is not really a new science.

 

[Mark911]

I finished the twin TB setup with some good and some bad results. I have a 90mm pre SC TB and a 75mm post TB using a 1-1 ratio opening. As suspected, the larger 90mm plate offers a larger cross section at every opening position and therefore provides a slight 0 to 2 lbs positive pressure against the post TB plate during part throttle application. The SC noise is very reasonable at this level and response is GREAT! Occasionally under mild load the SC can produce significant pressure (3-4 lbs) against a partially open post TB plate. This pressure pushes past the post plate and reduces manifold vacuum enough to actually close the internal bypass valve and manifold pressure quickly increases to 8-9 lbs. It feels like a turbo spooling up, pretty wild. The bad side is that there's simply too much air being bypassed (recirculated) through the SCer at cruising speed. This naturally raises the air temps which increases the heat burden on the intercooler. ultimately, the equilibrium water temp in the IC system is about 10-15 degrees F warmer then before (about 110 degrees on a 85 degree day at freeway speed). That equates to a no load IAT of about 125 degrees F, too warm in my book as it's only going to get hotter as the load increases. Not to mention that the SCer gets pretty darn hot.

I've been trying to figure out a way to better modulate the two TBs to minimize this conditions. Altering the throttle cable cam profiles seems like it would take months of trail and error. Going to a fly by wire type TB for the pre SC might work with some programming, but I'd prefer to stay away from more electronics. My final thought was to come up with a configuration where I could dump ALL the air coming from the SC to atmosphere (before the IC) during low load while allowing the post TB to perform as usual. This would require a third flow direction control valve (TB) to re-direct air from the SC to a dump valve (forth TB) and then to atmosphere. Both valves would be controlled by manifold vacuum, the directional valve being normally open (diverting air at load loads but allowing flow through the IC at load) and the dump valve normally closed (allowing air to vent to atmosphere at low loads but sealing the SC to IC path from boost leakage during load). Finally, a fifth valve positional directly AFTER directional control vale would be needed to allow the engine to breath normally when the directional valve is in the closed (diverting air) position. This again would be a manifold vacuum controlled normally closed TB (allowing air to move into the engine through the IC at low loads but again sealing the SC to IC path from boost leakage during load). obviously, an air cleaner would need to be integrated onto the end of this control valve.

Since the directional control TB would see full load flow, it would need to be sizeable, in my case 90mm. The other two valve could be considerably smaller, possibly 50mm or so. To avoid noise issues, a simple muffler could be installed onto the end of the dump valve.

This arrangement, although seemingly complex, solves many problems. The SCer will still try and produce boost (even though the pre TB and bypass are still limiting it), but any flow will simply exit the dump valve. This should actually COOL the SCer and reduce air temps and drive HP at low load. Since there is NO hot air passing through the IC at low loads, low load heat soak of the IC and IC water will be a thing of the past. Finally, breathing ambient air through the IC during low load operation could actually cool it down, making it more efficient when needed. And of course, all three of the new valves can be easily adjusted using the manifold vacuum diaphragm spring to close/open so that none of the GOOD aspects of the twin TB configuration are lost.

Just a few thought. Mark, mscperformance.com